Internet Draft






Internet Engineering Task Force                                   SIP WG
Internet Draft                    Handley/Schulzrinne/Schooler/Rosenberg
draft-ietf-sip-rfc2543bis-00.txt   ACIRI/Columbia U./Caltech/dynamicsoft
July 13, 2000
Expires: December 2000


                    SIP: Session Initiation Protocol

STATUS OF THIS MEMO

   This document is an Internet-Draft and is in full conformance with
   all provisions of Section 10 of RFC2026.

   Internet-Drafts are working documents of the Internet Engineering
   Task Force (IETF), its areas, and its working groups.  Note that
   other groups may also distribute working documents as Internet-
   Drafts.

   Internet-Drafts are draft documents valid for a maximum of six months
   and may be updated, replaced, or obsoleted by other documents at any
   time.  It is inappropriate to use Internet-Drafts as reference
   material or to cite them other than as "work in progress".

     The list of current Internet-Drafts can be accessed at
     http://www.ietf.org/ietf/1id-abstracts.txt

     The list of Internet-Draft Shadow Directories can be accessed at
     http://www.ietf.org/shadow.html.
Abstract

   The Session Initiation Protocol (SIP) is an application-layer control
   (signaling) protocol for creating, modifying and terminating sessions
   with one or more participants. These sessions include Internet
   multimedia conferences, Internet telephone calls and multimedia
   distribution. Members in a session can communicate via multicast or
   via a mesh of unicast relations, or a combination of these.

   SIP invitations used to create sessions carry session descriptions
   which allow participants to agree on a set of compatible media types.
   SIP supports user mobility by proxying and redirecting requests to
   the user's current location. Users can register their current
   location.  SIP is not tied to any particular conference control
   protocol. SIP is designed to be independent of the lower-layer
   transport protocol and can be extended with additional capabilities.


1 Introduction

1.1 Overview of SIP Functionality



Handley/Schulzrinne/Schooler/Rosenberg                        [Page 1]

Internet Draft                    SIP                      July 13, 2000


   The Session Initiation Protocol (SIP) is an application-layer control
   protocol that can establish, modify and terminate multimedia sessions
   or calls. These multimedia sessions include multimedia conferences,
   distance learning, Internet telephony and similar applications. SIP
   can invite both persons and "robots", such as a media storage
   service.  SIP can invite parties to both unicast and multicast
   sessions; the initiator does not necessarily have to be a member of
   the session to which it is inviting. Media and participants can be
   added to an existing session.

   SIP can be used to initiate sessions as well as invite members to
   sessions that have been advertised and established by other means.
   Sessions can be advertised using multicast protocols such as SAP,
   electronic mail, news groups, web pages or directories (LDAP), among
   others.

   SIP transparently supports name mapping and redirection services,
   allowing the implementation of ISDN and Intelligent Network telephony
   subscriber services. These facilities also enable personal mobility.
   In the parlance of telecommunications intelligent network services,
   this is defined as: "Personal mobility is the ability of end users to
   originate and receive calls and access subscribed telecommunication
   services on any terminal in any location, and the ability of the
   network to identify end users as they move. Personal mobility is
   based on the use of a unique personal identity (i.e., personal
   number)." [1]. Personal mobility complements terminal mobility, i.e.,
   the ability to maintain communications when moving a single end
   system from one subnet to another.

   SIP supports five facets of establishing and terminating multimedia
   communications:

        User location: determination of the end system to be used for
             communication;

        User capabilities: determination of the media and media
             parameters to be used;

        User availability: determination of the willingness of the
             called party to engage in communications;

        Call setup: "ringing", establishment of call parameters at both
             called and calling party;

        Call handling: including transfer and termination of calls.

   SIP can also initiate multi-party calls using a multipoint control
   unit (MCU) or fully-meshed interconnection instead of multicast.



Handley/Schulzrinne/Schooler/Rosenberg                        [Page 2]

Internet Draft                    SIP                      July 13, 2000


   Internet telephony gateways that connect Public Switched Telephone
   Network (PSTN) parties can also use SIP to set up calls between them.

   SIP is designed as part of the overall IETF multimedia data and
   control architecture currently incorporating protocols such as RSVP
   (RFC 2205 [2]) for reserving network resources, the real-time
   transport protocol (RTP) (RFC 1889 [3]) for transporting real-time
   data and providing QOS feedback, the real-time streaming protocol
   (RTSP) (RFC 2326 [4]) for controlling delivery of streaming media,
   the session announcement protocol (SAP) [5] for advertising
   multimedia sessions via multicast and the session description
   protocol (SDP) (RFC 2327 [6]) for describing multimedia sessions.
   However, the functionality and operation of SIP does not depend on
   any of these protocols.

   SIP can also be used in conjunction with other call setup and
   signaling protocols. In that mode, an end system uses SIP exchanges
   to determine the appropriate end system address and protocol from a
   given address that is protocol-independent. For example, SIP could be
   used to determine that the party can be reached via H.323 [7], obtain
   the H.245 [8] gateway and user address and then use H.225.0 [9] to
   establish the call.

   In another example, SIP might be used to determine that the callee is
   reachable via the PSTN and indicate the phone number to be called,
   possibly suggesting an Internet-to-PSTN gateway to be used.

   SIP does not offer conference control services such as floor control
   or voting and does not prescribe how a conference is to be managed,
   but SIP can be used to introduce conference control protocols. SIP
   does not allocate multicast addresses.

   SIP can invite users to sessions with and without resource
   reservation.  SIP does not reserve resources, but can convey to the
   invited system the information necessary to do this.

1.2 Terminology

   In this document, the key words "MUST", "MUST NOT", "REQUIRED",
   "SHALL", "SHALL NOT", "SHOULD", "SHOULD NOT", "RECOMMENDED", "MAY",
   and "OPTIONAL" are to be interpreted as described in RFC 2119 [10]
   and indicate requirement levels for compliant SIP implementations.

1.3 Definitions

   This specification uses a number of terms to refer to the roles
   played by participants in SIP communications. The definitions of
   client, server and proxy are similar to those used by the Hypertext



Handley/Schulzrinne/Schooler/Rosenberg                        [Page 3]

Internet Draft                    SIP                      July 13, 2000


   Transport Protocol (HTTP) (RFC 2616 [11]). The terms and generic
   syntax of URI and URL are defined in RFC 2396 [12]. The following
   terms have special significance for SIP.

        Call: A call consists of all participants in a conference
             invited by a common source. A SIP call is identified by a
             globally unique call-id (Section 6.13). Thus, if a user is,
             for example, invited to the same multicast session by
             several people, each of these invitations will be a unique
             call. A point-to-point Internet telephony conversation maps
             into a single SIP call. In a multiparty conference unit
             (MCU) based call-in conference, each participant uses a
             separate call to invite himself to the MCU.

        Call leg:  A call leg is identified by the combination of the
             Call-ID header field and the addr-spec and tag of the To
             and From header fields. Only the user and hostport parts of
             the addr-spec are significant.  Within the same Call-ID,
             requests with From A and To value B belong to the same call
             leg as the requests in the opposite direction, i.e., From B
             and To A.

        Client: An application program that sends SIP requests. Clients
             may or may not interact directly with a human user. User
             agents and proxies contain clients (and servers).

        Conference: A multimedia session (see below), identified by a
             common session description. A conference can have zero or
             more members and includes the cases of a multicast
             conference, a full-mesh conference and a two-party
             "telephone call", as well as combinations of these.  Any
             number of calls can be used to create a conference.

        Downstream: Requests sent in the direction from the caller to
             the callee (i.e., user agent client to user agent server).

        Final response: A response that terminates a SIP transaction, as
             opposed to a provisional response that does not. All 2xx,
             3xx, 4xx, 5xx and 6xx responses are final.

        Initiator, calling party, caller: The party initiating a
             conference invitation. Note that the calling party does not
             have to be the same as the one creating the conference.

        Invitation: A request sent to a user (or service) requesting
             participation in a session. A successful SIP invitation
             consists of two transactions: an INVITE request followed by
             an ACK request.



Handley/Schulzrinne/Schooler/Rosenberg                        [Page 4]

Internet Draft                    SIP                      July 13, 2000


        Invitee, invited user, called party, callee: The person or
             service that the calling party is trying to invite to a
             conference.

        Isomorphic request or response: Two requests or responses are
             defined to be isomorphic for the purposes of this document
             if they have the same values for the Call-ID, To, From and
             CSeq header fields. In addition, isomorphic requests have
             to have the same Request-URI.

        Location server: See location service.

        Location service: A location service is used by a SIP redirect
             or proxy server to obtain information about a callee's
             possible location(s). Examples of sources of location
             information include SIP registrars, databases or mobility
             registration protocols. Location services are offered by
             location servers. Location servers MAY be part of a SIP
             server, but the manner in which a SIP server requests
             location services is beyond the scope of this document.

        Outbound proxy: A proxy that is located near the originator of
             requests. It receives all outgoing requests from a
             particular UAC, including those requests whose Request-URLs
             identify a host other than the outbound proxy. The outbound
             proxy sends these requests, after any local processing, to
             the address indicated in the request-URI. (All other proxy
             servers are simply referred as proxies, not inbound
             proxies.)

        Parallel search: In a parallel search, a proxy issues several
             requests to possible user locations upon receiving an
             incoming request.  Rather than issuing one request and then
             waiting for the final response before issuing the next
             request as in a sequential search , a parallel search
             issues requests without waiting for the result of previous
             requests.

        Provisional response: A response used by the server to indicate
             progress, but that does not terminate a SIP transaction.
             1xx responses are provisional, other responses are
             considered final.

        Proxy, proxy server: An intermediary program that acts as both a
             server and a client for the purpose of making requests on
             behalf of other clients. Requests are serviced internally
             or by passing them on, possibly after translation, to other
             servers. A proxy interprets, and, if necessary, rewrites a



Handley/Schulzrinne/Schooler/Rosenberg                        [Page 5]

Internet Draft                    SIP                      July 13, 2000


             request message before forwarding it.

        Redirect server: A redirect server is a server that accepts a
             SIP request, maps the address into zero or more new
             addresses and returns these addresses to the client. Unlike
             a proxy server , it does not initiate its own SIP request.
             Unlike a user agent server , it does not accept calls.

        Registrar: A registrar is a server that accepts REGISTER
             requests. A registrar is typically co-located with a proxy
             or redirect server and MAY make its information available
             through the location server.

        Ringback: Ringback is the signaling tone produced by the calling
             client's application indicating that a called party is
             being alerted (ringing).

        Server: A server is an application program that accepts requests
             in order to service requests and sends back responses to
             those requests.  Servers are either proxy, redirect or user
             agent servers or registrars.

        Session: From the SDP specification: "A multimedia session is a
             set of multimedia senders and receivers and the data
             streams flowing from senders to receivers. A multimedia
             conference is an example of a multimedia session." (RFC
             2327 [6]) (A session as defined for SDP can comprise one or
             more RTP sessions.) As defined, a callee can be invited
             several times, by different calls, to the same session. If
             SDP is used, a session is defined by the concatenation of
             the user name , session id , network type , address type
             and address elements in the origin field.

        (SIP) transaction: A SIP transaction occurs between a client and
             a server and comprises all messages from the first request
             sent from the client to the server up to a final (non-1xx)
             response sent from the server to the client. A transaction
             is identified by the CSeq sequence number (Section 6.21)
             within a single call leg.  The ACK request has the same
             CSeq number as the corresponding INVITE request, but
             comprises a transaction of its own.

        Upstream: Responses sent in the direction from the user agent
             server to the user agent client.

        URL-encoded: A character string encoded according to RFC 1738,
             Section 2.2 [13].




Handley/Schulzrinne/Schooler/Rosenberg                        [Page 6]

Internet Draft                    SIP                      July 13, 2000


        User agent client (UAC): A user agent client is a client
             application that initiates a SIP request.

        User agent server (UAS): A user agent server is a server
             application that contacts the user when a SIP request is
             received and that returns a response on behalf of the user.
             The response accepts, rejects or redirects the request.

        User agent (UA): An application which can act both as a user
             agent client and user agent server.

   An application program MAY be capable of acting both as a client and
   a server. For example, a typical multimedia conference control
   application would act as a user agent client to initiate calls or to
   invite others to conferences and as a user agent server to accept
   invitations.  The role of UAC and UAS as well as proxy and redirect
   servers are defined on a request-by-request basis. For example, the
   user agent initiating a call acts as a UAC when sending the initial
   INVITE request and as a UAS when receiving a BYE request from the
   callee.  Similarly, the same software can act as a proxy server for
   one request and as a redirect server for the next request.

   Proxy, redirect, location and registrar servers defined above are
   logical entities; implementations MAY combine them into a single
   application program.  The properties of the different SIP server
   types are summarized in Table 1.


    property                   redirect  proxy   user agent  registrar
                                server   server    server
    __________________________________________________________________
    also acts as a SIP client     no      yes        no         no
    returns 1xx status           yes      yes       yes         yes
    returns 2xx status            no      yes       yes         yes
    returns 3xx status           yes      yes       yes         yes
    returns 4xx status           yes      yes       yes         yes
    returns 5xx status           yes      yes       yes         yes
    returns 6xx status            no      yes       yes         yes
    inserts Via header            no      yes        no         no
    accepts ACK                  yes      yes       yes         no


   Table 1: Properties of the different SIP server types


1.4 Overview of SIP Operation

   This section explains the basic protocol functionality and operation.
   Callers and callees are identified by SIP addresses, described in


Handley/Schulzrinne/Schooler/Rosenberg                        [Page 7]

Internet Draft                    SIP                      July 13, 2000


   Section 1.4.1. When making a SIP call, a caller first locates the
   appropriate server (Section 1.4.2) and then sends a SIP request
   (Section 1.4.3). The most common SIP operation is the invitation
   (Section 1.4.4). Instead of directly reaching the intended callee, a
   SIP request may be redirected or may trigger a chain of new SIP
   requests by proxies (Section 1.4.5). Users can register their
   location(s) with SIP servers (Section 4.2.6).

1.4.1 SIP Addressing

   The "objects" addressed by SIP are users at hosts, identified by a
   SIP URL. The SIP URL takes a form similar to a mailto or telnet URL,
   i.e., user@host.  The user part is a user name or a telephone number.
   The host part is either a domain name or a numeric network address.
   See section 2 for a detailed discussion of SIP URL's.

   A user's SIP address can be obtained out-of-band, can be learned via
   existing media agents, can be included in some mailers' message
   headers, or can be recorded during previous invitation interactions.
   In many cases, a user's SIP URL can be guessed from their email
   address.

   A SIP URL address can designate an individual (possibly located at
   one of several end systems), the first available person from a group
   of individuals or a whole group. The form of the address, for
   example, sip:sales@example.com , is not sufficient, in general, to
   determine the intent of the caller.

   If a user or service chooses to be reachable at an address that is
   guessable from the person's name and organizational affiliation, the
   traditional method of ensuring privacy by having an unlisted "phone"
   number is compromised. However, unlike traditional telephony, SIP
   offers authentication and access control mechanisms and can avail
   itself of lower-layer security mechanisms, so that client software
   can reject unauthorized or undesired call attempts.

1.4.2 Locating a SIP Server

   When a client wishes to send a request, the client either sends it to
   a locally configured SIP proxy server (as in HTTP), independent of
   the Request-URI, or sends it to the IP address and port corresponding
   to the Request-URI.

   For the latter case, the client must determine the protocol, port and
   IP address of a server to which to send the request. A client SHOULD
   follow the steps below to obtain this information,

   At each step, unless stated otherwise, the client SHOULD try to



Handley/Schulzrinne/Schooler/Rosenberg                        [Page 8]

Internet Draft                    SIP                      July 13, 2000


   contact a server at the port number listed in the Request-URI. If no
   port number is present in the Request-URI, the client uses port 5060.
   If the Request-URI specifies a protocol (TCP or UDP), the client
   contacts the server using that protocol. If no protocol is specified,
   the client tries UDP (if UDP is supported). If the attempt fails, or
   if the client doesn't support UDP but supports TCP, it then tries
   TCP.

   A client SHOULD be able to interpret explicit network notifications
   (such as ICMP messages) which indicate that a server is not
   reachable, rather than relying solely on timeouts. (For socket-based
   programs:  For TCP, connect() returns ECONNREFUSED if the client
   could not connect to a server at that address. For UDP, the socket
   needs to be bound to the destination address using connect() rather
   than sendto() or similar so that a second write() or send() fails
   with ECONNREFUSED if there is no server listening) If the client
   finds the server is not reachable at a particular address, it SHOULD
   behave as if it had received a 400-class error response to that
   request.

   The client tries to find one or more addresses for the SIP server by
   querying DNS. If a step elicits no addresses, the client continues to
   the next step. However if a step elicits one or more addresses, but
   no SIP server at any of those addresses responds, then the client
   concludes the server is down and doesn't continue on to the next
   step.

   The service identifier for DNS SRV records [14] is "_sip".  SRV
   records contain port numbers for servers, in addition to IP
   addresses; the client always uses this port number when contacting
   the SIP server. Otherwise, the port number in the SIP URI is used, if
   present. If there is no port number in the URI, the default port,
   5060, is used.

   The procedure is described below.

        1.   If the maddr parameter of the Request-URI is an IP address,
             the client contacts the server at the given address and
             ignores the remaining steps.

        2.   If the maddr parameter of the Request-URI is a host name,
             the client proceeds to the last step.

        3.   If the host portion of the Request-URI is an IP address,
             the client contacts the server at the given address.
             Otherwise, the client proceeds to the next step.

        4.   The Request-URI is examined. If it contains an explicit



Handley/Schulzrinne/Schooler/Rosenberg                        [Page 9]

Internet Draft                    SIP                      July 13, 2000


             port number other than 5060, the next two steps are
             skipped.

        5.   The Request-URI is examined. If it does not specify a
             protocol (TCP or UDP), the client queries the name server
             for SRV records for both UDP (if supported by the client)
             and TCP (if supported by the client) SIP servers. The
             format of these queries is defined in RFC 2782 [14]. The
             results of the query or queries are merged together and
             ordered based on priority. Then, the searching technique
             outlined in RFC 2782 [14] is used to select servers in
             order.  If DNS does not return any records, the user goes
             to the last step.  Otherwise, the user attempts to contact
             each server in the order listed.  If none of the server can
             be contacted, the user gives up.

        6.   If the Request-URI specifies a protocol (TCP or UDP) that
             is supported by the client, the client queries the name
             server for SRV records for SIP servers of that protocol
             type only. If the client does not support the protocol
             specified in the Request-URI, it gives up. The searching
             technique outlined in RFC 2782 [14] is used to select
             servers from the DNS response in order. If DNS does not
             return any records, the user goes to the last step.
             Otherwise, the user attempts to contact each server in the
             order listed. If no server is contacted, the user gives up.

        7.   The client queries the DNS server for address records for
             the host portion of the maddr parameter, if preent, or the
             Request-URI. If the DNS server returns no address records,
             the client stops, as it has been unable to locate a server.
             Address records include A RR's, AAAA RR's, or other similar
             records, chosen according to the client's network protocol
             capabilities.


        There are no mandatory rules on how to select a host name
        for a SIP server. Users are encouraged to name their SIP
        servers using the sip.domainname (e.g., sip.example.com )
        convention, as specified in RFC 2219 [15]. Users may only
        know an email address instead of a full SIP URL for a
        callee, however. In that case, implementations may be able
        to increase the likelihood of reaching a SIP server for
        that domain by constructing a SIP URL from that email
        address by prefixing the host name with "sip.". In the
        future, this mechanism is likely to become unnecessary as
        SRV records, described above, become widely available.




Handley/Schulzrinne/Schooler/Rosenberg                       [Page 10]

Internet Draft                    SIP                      July 13, 2000


   A client MAY cache the list of DNS query results if one of the
   addresses was contacted successfully.  Request for the same
   transaction SHOULD be sent to the same network address. Other
   requests from the same client select a server from the list of
   addresses cached, using the SRV load-balancing mechanism if
   applicable. The client must invalidate this list and retry the DNS
   query according to the rules in RFC1035 [16].

   A client MAY omit attempting to reach a server which it had failed to
   reach for a previous request.

        If the DNS time-to-live value exceeds a few minutes,
        servers generating a large number of requests are probably
        well advised to retry failed servers every few minutes.

1.4.3 SIP Transaction

   Once the host part has been resolved to a SIP server, the client
   sends one or more SIP requests to that server and receives one or
   more responses from the server. A request (and its retransmissions)
   together with the responses triggered by that request make up a SIP
   transaction.  All responses to a request contain the same values in
   the Call-ID, CSeq, To, and From fields (with the possible addition of
   a tag in the To field (section 6.44)). This allows responses to be
   matched with requests. The ACK request following an INVITE is not
   part of the transaction since it may traverse a different set of
   hosts.

   If TCP is used, request and responses within a single SIP transaction
   are carried over the same TCP connection (see Section 10). Several
   SIP requests from the same client to the same server MAY use the same
   TCP connection or MAY use a new connection for each request.

   If the client sent the request via unicast UDP, the response is sent
   to the address contained in the next Via header field (Section 6.47)
   of the response. If the request is sent via multicast UDP, the
   response is directed to the same multicast address and destination
   port. For UDP, reliability is achieved using retransmission (Section
   10).

   The SIP message format and operation is independent of the transport
   protocol.

1.4.4 SIP Invitation

   A successful SIP invitation consists of two requests, INVITE followed
   by ACK. The INVITE (Section 4.2.1) request asks the callee to join a
   particular conference or establish a two-party conversation. After



Handley/Schulzrinne/Schooler/Rosenberg                       [Page 11]

Internet Draft                    SIP                      July 13, 2000


   the callee has agreed to participate in the call, the caller confirms
   that it has received that response by sending an ACK (Section 4.2.2)
   request. If the caller no longer wants to participate in the call, it
   sends a BYE request instead of an ACK.

   The INVITE request typically contains a session description, for
   example written in SDP (RFC 2327 [6]) format, that provides the
   called party with enough information to join the session. For
   multicast sessions, the session description enumerates the media
   types and formats that are allowed to be distributed to that session.
   For a unicast session, the session description enumerates the media
   types and formats that the caller is willing to use and where it
   wishes the media data to be sent. In either case, if the callee
   wishes to accept the call, it responds to the invitation by returning
   a similar description listing the media it wishes to use. For a
   multicast session, the callee SHOULD only return a session
   description if it is unable to receive the media indicated in the
   caller's description or wants to receive data via unicast.

   The protocol exchanges for the INVITE method are shown in Fig. 1 for
   a proxy server and in Fig. 2 for a redirect server. (Note that the
   messages shown in the figures have been abbreviated slightly.) In
   Fig. 1, the proxy server accepts the INVITE request (step 1),
   contacts the location service with all or parts of the address (step
   2) and obtains a more precise location (step 3). The proxy server
   then issues a SIP INVITE request to the address(es) returned by the
   location service (step 4). The user agent server alerts the user
   (step 5) and returns a success indication to the proxy server (step
   6). The proxy server then returns the success result to the original
   caller (step 7). The receipt of this message is confirmed by the
   caller using an ACK request, which is forwarded to the callee (steps
   8 and 9). Note that an ACK can also be sent directly to the callee,
   bypassing the proxy. All requests and responses have the same Call-
   ID.


   The redirect server shown in Fig. 2 accepts the INVITE request (step
   1), contacts the location service as before (steps 2 and 3) and,
   instead of contacting the newly found address itself, returns the
   address to the caller (step 4), which is then acknowledged via an ACK
   request (step 5). The caller issues a new request, with the same
   call-ID but a higher CSeq, to the address returned by the first
   server (step 6). In the example, the call succeeds (step 7). The
   caller and callee complete the handshake with an ACK (step 8).


   The next section discusses what happens if the location service
   returns more than one possible alternative.



Handley/Schulzrinne/Schooler/Rosenberg                       [Page 12]

Internet Draft                    SIP                      July 13, 2000






                                         +....... cs.columbia.edu .......+
                                         :                               :
                                         : (~~~~~~~~~~)                  :
                                         : ( location )                  :
                                         : ( service  )                  :
                                         : (~~~~~~~~~~)                  :
                                         :     ^    |                    :
                                         :     | hgs@lab                 :
                                         :    2|   3|                    :
                                         :     |    |                    :
                                         : henning  |                    : 
+.. cs.tu-berlin.de ..+ 1: INVITE        :     |    |                    :
:                     :    henning@cs.col:     |   \/ 4: INVITE  5: ring :
: cz@cs.tu-berlin.de ========================>(~~~~~~)=========>(~~~~~~) :
:                    <........................(      )<.........(      ) :
:                     : 7: 200 OK        :    (      )6: 200 OK (      ) :
:                     :                  :    ( work )          ( lab  ) :
:                     : 8: ACK           :    (      )9: ACK    (      ) :
:                    ========================>(~~~~~~)=========>(~~~~~~) :
+.....................+                  +...............................+

  ====> SIP request                                                         
  ....> SIP response                                                       
  
   ^
   |    non-SIP protocols                                                  
   |
  

   Figure 1: Example of SIP proxy server


1.4.5 Locating a User

   A callee may move between a number of different end systems over
   time.  These locations can be dynamically registered with the SIP
   server (Sections 1.4.7, 4.2.6). A location server MAY also use one or
   more other protocols, such as finger (RFC 1288 [17]), rwhois (RFC
   2167 [18]), LDAP (RFC 1777 [19]), multicast-based protocols [20] or
   operating-system dependent mechanisms to actively determine the end
   system where a user might be reachable. A location server MAY return
   several locations because the user is logged in at several hosts
   simultaneously or because the location server has (temporarily)
   inaccurate information. The SIP server combines the results to yield



Handley/Schulzrinne/Schooler/Rosenberg                       [Page 13]

Internet Draft                    SIP                      July 13, 2000






                                         +....... cs.columbia.edu .......+
                                         :                               :
                                         : (~~~~~~~~~~)                  :
                                         : ( location )                  :
                                         : ( service  )                  :
                                         : (~~~~~~~~~~)                  :
                                         :    ^   |                      :
                                         :    | hgs@lab                  :
                                         :   2|  3|                      :
                                         :    |   |                      :
                                         : henning|                      : 
+.. cs.tu-berlin.de ..+ 1: INVITE        :    |   |                      :
:                     :    henning@cs.col:    |   \/                     : 
: cz@cs.tu-berlin.de =======================>(~~~~~~)                    : 
:       | ^ |        <.......................(      )                    :
:       | . |         : 4: 302 Moved     :   (      )                    :
:       | . |         :    hgs@lab       :   ( work )                    :
:       | . |         :                  :   (      )                    :
:       | . |         : 5: ACK           :   (      )                    :
:       | . |        =======================>(~~~~~~)                    :
:       | . |         :                  :                               :
+.......|...|.........+                  :                               :
        | . |                            :                               :
        | . |                            :                               :
        | . |                            :                               :
        | . |                            :                               :
        | . | 6: INVITE hgs@lab.cs.columbia.edu                 (~~~~~~) : 
        | . ==================================================> (      ) :
        | ..................................................... (      ) :
        |     7: 200 OK                  :                      ( lab  ) : 
        |                                :                      (      ) :
        |     8: ACK                     :                      (      ) :
        ======================================================> (~~~~~~) :
                                         +...............................+ 
                                                                          
  ====> SIP request                                                        
  ....> SIP response                                                       
    
    ^
    |   non-SIP protocols                                                  
    |




   Figure 2: Example of SIP redirect server

Handley/Schulzrinne/Schooler/Rosenberg                       [Page 14]

Internet Draft                    SIP                      July 13, 2000


   a list of a zero or more locations.

   The action taken on receiving a list of locations varies with the
   type of SIP server. A SIP redirect server returns the list to the
   client as Contact headers (Section 6.15). A SIP proxy server can
   sequentially or in parallel try the addresses until the call is
   successful (2xx response) or the callee has declined the call (6xx
   response). With sequential attempts, a proxy server can implement an
   "anycast" service.

   If a proxy server forwards a SIP request, it MUST add itself to the
   beginning of the list of forwarders noted in the Via (Section 6.47)
   headers. The Via trace ensures that replies can take the same path
   back, ensuring correct operation through compliant firewalls and
   avoiding request loops. On the response path, each host MUST remove
   its Via, so that routing internal information is hidden from the
   callee and outside networks. A proxy server MUST check that it does
   not generate a request to a host listed in the Via sent-by, via-
   received or via-maddr parameters (Section 6.47). (Note: If a host has
   several names or network addresses, this does not always work.  Thus,
   each host also checks if it is part of the Via list.)

   A SIP invitation may traverse more than one SIP proxy server. If one
   of these "forks" the request, i.e., issues more than one request in
   response to receiving the invitation request, it is possible that a
   client is reached, independently, by more than one copy of the
   invitation request. Each of these copies bears the same Call-ID. The
   user agent MUST return the same status response returned in the first
   response. Duplicate requests are not an error.

1.4.6 Changing an Existing Session

   In some circumstances, it is desirable to change the parameters of an
   existing session. This is done by re-issuing the INVITE, using the
   same Call-ID, but a new or different body or header fields to convey
   the new information. This re INVITE MUST have a higher CSeq than any
   previous request from the client to the server.

   For example, two parties may have been conversing and then want to
   add a third party, switching to multicast for efficiency.  One of the
   participants invites the third party with the new multicast address
   and simultaneously sends an INVITE to the second party, with the new
   multicast session description, but with the old call identifier.

1.4.7 Registration Services

   The REGISTER request allows a client to let a proxy or redirect
   server know at which address(es) it can be reached. A client MAY also



Handley/Schulzrinne/Schooler/Rosenberg                       [Page 15]

Internet Draft                    SIP                      July 13, 2000


   use it to install call handling features at the server.

1.5 Protocol Properties

1.5.1 Minimal State

   A single conference session or call involves one or more SIP
   request-response transactions. Proxy servers do not have to keep
   state for a particular call, however, they MAY maintain state for a
   single SIP transaction, as discussed in Section 12. For efficiency, a
   server MAY cache the results of location service requests.

1.5.2 Lower-Layer-Protocol Neutral

   SIP makes minimal assumptions about the underlying transport and
   network-layer protocols. The lower-layer can provide either a packet
   or a byte stream service, with reliable or unreliable service.

   In an Internet context, SIP is able to utilize both UDP and TCP as
   transport protocols, among others. UDP allows the application to more
   carefully control the timing of messages and their retransmission, to
   perform parallel searches without requiring TCP connection state for
   each outstanding request, and to use multicast. Routers can more
   readily snoop SIP UDP packets. TCP allows easier passage through
   existing firewalls.

   When TCP is used, SIP can use one or more connections to attempt to
   contact a user or to modify parameters of an existing conference.
   Different SIP requests for the same SIP call MAY use different TCP
   connections or a single persistent connection, as appropriate.

   For concreteness, this document will only refer to Internet
   protocols.  However, SIP MAY also be used directly with protocols
   such as ATM AAL5, IPX, frame relay or X.25. The necessary naming
   conventions are beyond the scope of this document. User agents SHOULD
   implement both UDP and TCP transport. Proxy, registrar, and redirect
   servers MUST implement both UDP and TCP transport.

1.5.3 Text-Based

   SIP is text-based, using ISO 10646 in UTF-8 encoding throughout. This
   allows easy implementation in languages such as Java, Tcl and Perl,
   allows easy debugging, and most importantly, makes SIP flexible and
   extensible. As SIP is used for initiating multimedia conferences
   rather than delivering media data, it is believed that the additional
   overhead of using a text-based protocol is not significant.

2 SIP Uniform Resource Locators



Handley/Schulzrinne/Schooler/Rosenberg                       [Page 16]

Internet Draft                    SIP                      July 13, 2000


   SIP URLs are used within SIP messages to indicate the originator
   (From), current destination (Request-URI) and final recipient (To) of
   a SIP request, and to specify redirection addresses (Contact). A SIP
   URL can also be embedded in web pages or other hyperlinks to indicate
   that a particular user or service can be called via SIP. When used as
   a hyperlink, the SIP URL indicates the use of the INVITE method.

   The SIP URL scheme is defined to allow setting SIP request-header
   fields and the SIP message-body.


        This corresponds to the use of mailto: URLs. It makes it
        possible, for example, to specify the subject, urgency or
        media types of calls initiated through a web page or as
        part of an email message.

   A SIP URL follows the guidelines of RFC 2396 [12] and has the syntax
   shown in Fig. 3. The syntax is described using Augmented Backus-Naur
   Form (see Section C). Note that reserved characters have to be
   escaped and that the "set of characters reserved within any given URI
   component is defined by that component. In general, a character is
   reserved if the semantics of the URI changes if the character is
   replaced with its escaped US-ASCII encoding" [12].  Excluded US-ASCII
   characters [12], such as space and control characters and characters
   used as URL delimiters, also MUST be escaped.


   The URI character classes referenced above are described in Appendix
   C.

   The components of the SIP URI have the following meanings.

        user: The name of the user addressed. Note that this field MAY
             be empty where the destination host does not have a notion
             of users, e.g., for embedded devices.

        telephone-subscriber: If the host is an Internet telephony
             gateway, a telephone-subscriber field MAY be used instead
             of a user field. The telephone-subscriber field uses the
             notation of RFC 2806 [21]. Any characters of the un-escaped
             "telephone-subscriber" that are not either in the set
             "unreserved" or "user-unreserved" MUST be escaped. The set
             of characters not reserved in the RFC 2806 description of
             telephone-subscriber contains a number of characters in
             various syntax elements that need to be escaped when used
             in SIP URLs, for example quotation marks (%x22), hash
             (%x23), colon (%x3a), at-sign (%x40) and the "unwise"
             characters, i.e., punctuation of %x5b and above.



Handley/Schulzrinne/Schooler/Rosenberg                       [Page 17]

Internet Draft                    SIP                      July 13, 2000


             The telephone number is a special case of a user name and
             cannot be distinguished by a BNF. Thus, a URL parameter,
             user, is added to distinguish telephone numbers from user
             names.

             The user parameter value "phone" indicates that the user
             part contains a telephone number. Even without this
             parameter, recipients of SIP URLs MAY interpret the pre-@
             part as a telephone number if local restrictions on the
             name space for user name allow it.

             The user parameter value "np-queried" indicates that the
             user part contains a telephone number and that the number
             reflects the result of a query to the local number
             portability database.

        password: The SIP scheme MAY use the format "user:password" in
             the userinfo field. The use of passwords in the userinfo is
             NOT RECOMMENDED, because the passing of authentication
             information in clear text (such as URIs) has proven to be a
             security risk in almost every case where it has been used.

        host: The host part SHOULD be a fully-qualified domain name or
             numeric IP address.

             The mailto: URL and RFC 822 email addresses require that
             numeric host addresses ("host numbers") are enclosed in
             square brackets (presumably, since host names might be
             numeric), while host numbers without brackets are used for
             all other URLs. The SIP URL requires the latter form,
             without brackets.

        port: The port number to send a request to. If not present, the
             procedures outlined in Section 1.4.2 are used to determine
             the port number to send a request to.

        URL parameters: SIP URLs can define specific parameters of the
             request. URL parameters are added after the host component
             and are separated by semi-colons. The transport parameter
             determines the the transport mechanism to be used for
             sending SIP requests and responses. SIP can use any network
             transport protocol; parameter names are defined for UDP
             [22], TCP [23], TLS [24], and SCTP.  UDP is to be assumed
             when no explicit transport parameter is included.  The
             maddr parameter provides the server address to be contacted
             for this user, overriding the address supplied in the host
             field. This address is typically a multicast address, but
             could also be the address of a backup server. The ttl



Handley/Schulzrinne/Schooler/Rosenberg                       [Page 18]

Internet Draft                    SIP                      July 13, 2000




  SIP-URL         = "sip:" [ userinfo "@" ] hostport
                    url-parameters [ headers ]
  userinfo        = [ user | telephone-subscriber ] [ ":" password ]
  user            = *( unreserved | escaped
                  | "&" | "=" | "+" | "$" | "," | ";" | "?" | "/" )
  password        = *( unreserved | escaped
                  | "&" | "=" | "+" | "$" | "," )
  hostport        = host [ ":" port ]
  host            = hostname | IPv4address | IPv6reference
  hostname        = *( domainlabel "." ) toplabel [ "." ]
  domainlabel     = alphanum | alphanum *( alphanum | "-" ) alphanum
  toplabel        = alpha | alpha *( alphanum | "-" ) alphanum
  IPv4address     = 1*3DIGIT "." 1*3DIGIT "." 1*3DIGIT "." 1*3DIGIT
  IPv6reference   = "[" IPv6address "]"
  IPv6address     = hexpart [ ":" IPv4address ]
  hexpart         = hexseq | hexseq "::" [ hexseq ] | "::" [ hexseq ]
  hexseq          = hex4 *( ":" hex4)
  hex4            = 1*4HEX
  port            = 1*DIGIT
  url-parameters  = *( ";" url-parameter )
  url-parameter   = transport-param | user-param | method-param
                  | ttl-param | maddr-param | other-param
  transport-param = "transport=" 
                    ( "udp" | "tcp" | "sctp" | "tls" | other-transport )
  other-transport = token
  ttl-param       = "ttl=" ttl
  ttl             = 1*3DIGIT       ; 0 to 255
  maddr-param     = "maddr=" host
  user-param      = "user=" ( "phone" | "ip" | other-user )
  other-user      = token
  method-param    = "method=" Method
  tag-param       = "tag=" UUID
  UUID            = 1*( HEX | "-" )
  other-param     = pname [ "=" pvalue ]
  pname           = 1*paramchar
  pvalue          = 1*paramchar
  paramchar       = param-reserved | unreserved | escaped
  param-reserved  = "[" | "]" | "/" | "?" | ":" | "@" | "&" | "+" | "$"
  headers         = "?" header *( "&" header )
  header          = hname "=" hvalue
  hname           = 1*( hnv-unreserved | unreserved | escaped )
  hvalue          = *( hnv-unreserved | unreserved | escaped )
  hnv-unreserved  = "[" | "]" | "/" | "?" | ":" | "@" | "+" | "$"


   Figure 3: SIP URL syntax



Handley/Schulzrinne/Schooler/Rosenberg                       [Page 19]

Internet Draft                    SIP                      July 13, 2000


             multicast address and the transport protocol is UDP. The
             user parameter was described above. For example, to specify
             to call j.doe@big.com using multicast to 239.255.255.1 with
             a ttl of 15, the following URL would be used:


               sip:j.doe@big.com;maddr=239.255.255.1;ttl=15



             The transport, maddr, and ttl parameters MUST NOT be used
             in the From and To header fields and the Request-URI; they
             are ignored if present.

        Headers: Headers of the SIP request can be defined with the "?"
             mechanism within a SIP URL. The special hname "body"
             indicates that the associated hvalue is the message-body of
             the SIP INVITE request. Headers MUST NOT be used in the
             From and To header fields and the Request-URI; they are
             ignored if present.  hname and hvalue are encodings of a
             SIP header name and value, respectively. All URL reserved
             characters in the header names and values MUST be escaped.

        Method: The method of the SIP request can be specified with the
             method parameter.  This parameter MUST NOT be used in the
             From and To header fields and the Request-URI; they are
             ignored if present.

   Table 2 summarizes where the components of the SIP URL can be used.
   Entries marked "m" are mandatory, those marked "o" are optional, and
   those marked "-" are not allowed. For optional elements, the second
   column indicates the default value if the element is not present.


   Examples of SIP URLs are:

     sip:j.doe@big.com
     sip:j.doe:secret@big.com;transport=tcp
     sip:j.doe@big.com?subject=project
     sip:+1-212-555-1212:1234@gateway.com;user=phone
     sip:1212@gateway.com
     sip:alice@10.1.2.3
     sip:alice@example.com
     sip:alice
     sip:alice@registrar.com;method=REGISTER






Handley/Schulzrinne/Schooler/Rosenberg                       [Page 20]

Internet Draft                    SIP                      July 13, 2000



                  default    Req.-URI  To  From  Contact  Rec.-Route  external
   user           --            o      o    o       o         o          o
   password       --            o      o    -       o         o          o
   host           mandatory     m      m    m       m         m          m
   port           5060          o      o    o       o         o          o
   user-param     ip            o      o    o       o         o          o
   method         INVITE        -      -    -       o         -          o
   maddr-param    --            o      -    -       o         m          o
   ttl-param      1             o      -    -       o         -          o
   transp.-param  udp           o      -    -       o         -          o
   other-param    o             o      o    o       o         o          o
   headers        --            -      -    -       o         -          o


   Table 2: Use and default values of URL components  for  SIP  headers,
   Request-URI and references

   Within a SIP message, URLs are used to indicate the source and
   intended destination of a request, redirection addresses and the
   current destination of a request. Normally all these fields will
   contain SIP URLs.

2.1 SIP URL Comparison

   SIP URLs are compared for equality according to the following rules:

        o Comparisons of scheme name ("sip"), domain names, parameter
          names and header names are case-insensitive, all other
          comparisons are case-sensitive.

        o The ordering of parameters and headers is not significant in
          comparing SIP URLs.

        o user, password, host, port and any url-parameter parameters of
          the URI must match. If a component is omitted, it matches
          based on its default value. (For example, otherwise equivalent
          URLs without a port specification and with port 5060 match.)

          Note that the two URLs example.com and example.com:5060 ,
          while considered equal, may not lead to the same server, as
          the former causes a DNS SRV lookup, while the latter only uses
          the A record.

        o Characters other than those in the "reserved" and "unsafe"
          sets (see RFC 2396 [12]) are equivalent to their ""%" HEX HEX"
          encoding.




Handley/Schulzrinne/Schooler/Rosenberg                       [Page 21]

Internet Draft                    SIP                      July 13, 2000


        o An IP address that is the result of a DNS lookup of a host
          name does not match that host name.

   Thus, the following URLs are equivalent:

   sip:juser@
   sip:juser@ExAmPlE.CoM;Transport=udp


   while

   SIP:JUSER@ExAmPlE.CoM;Transport=udp
   sip:juser@ExAmPlE.CoM;Transport=UDP


   are not.

2.2 Non-SIP URLs

   SIP header fields and the Request-URI MAY contain non-SIP URLs, with
   the exceptions noted below. As an example, if a call from a telephone
   is relayed to the Internet via SIP, the SIP From header field might
   contain a tel: URL [21].

   In the following locations, only SIP URLs are allowed:

        o Request-URI in a REGISTER request;

        o Contact header field in INVITE, OPTIONS and BYE and 2xx
          responses.

   Implementations MAY compare non-SIP URLs by treating them as generic
   URIs [12] or, alternatively, compare them byte-by-byte.

3 SIP Message Overview

   SIP is a text-based protocol and uses the ISO 10646 character set in
   UTF-8 encoding (RFC 2279 [25]). Senders MUST terminate lines with a
   CRLF, but receivers MUST also interpret CR and LF by themselves as
   line terminators.  Only the combinations CR CR, LF LF and CRLF CRLF
   terminate the message header. Implementations MUST only send CRLF
   CRLF.

        CR and LF instead of CRLF is for backwards-compatibility;
        their use is deprecated.

   Except for the above difference in character sets and line
   termination, much of the message syntax is and header fields are



Handley/Schulzrinne/Schooler/Rosenberg                       [Page 22]

Internet Draft                    SIP                      July 13, 2000


   identical to HTTP/1.1; rather than repeating the syntax and semantics
   here we use [HX.Y] to refer to Section X.Y of the current HTTP/1.1
   specification (RFC 2616 [11]). In addition, we describe SIP in both
   prose and an augmented Backus-Naur form (ABNF). See section C for an
   overview of ABNF.

   Note, however, that SIP is not an extension of HTTP.

   Unlike HTTP, SIP MAY use UDP. When sent over TCP or UDP, multiple SIP
   transactions can be carried in a single TCP connection or UDP
   datagram. UDP datagrams, including all headers, SHOULD NOT be larger
   than the path maximum transmission unit (MTU) if the MTU is known, or
   1500 bytes if the MTU is unknown.


        The MTU of 1500 bytes accommodates encapsulation within the
        "typical" ethernet MTU without IP fragmentation. Recent
        studies [26] indicate that an MTU of 1500 bytes is a
        reasonable assumption. The next lower common MTU values are
        1006 bytes for SLIP and 296 for low-delay PPP (RFC 1191
        [27]). Thus, another reasonable value would be a message
        size of 950 bytes, to accommodate packet headers within the
        SLIP MTU without fragmentation.

   A SIP message is either a request from a client to a server, or a
   response from a server to a client.



        SIP-message  =  Request | Response


   Both Request (section 4) and Response (section 5) messages use the
   generic-message format of RFC 822 [28] for transferring entities (the
   body of the message). Both types of messages consist of a start-line,
   one or more header fields (also known as "headers"), an empty line
   (i.e., a line with nothing preceding the carriage-return line-feed
   (CRLF)) indicating the end of the header fields, and an optional
   message-body. To avoid confusion with similar-named headers in HTTP,
   we refer to the headers describing the message body as entity
   headers. These components are described in detail in the upcoming
   sections.



        generic-message  =  start-line
                            *message-header
                            CRLF



Handley/Schulzrinne/Schooler/Rosenberg                       [Page 23]

Internet Draft                    SIP                      July 13, 2000


                            [ message-body ]

        start-line       =  Request-Line |     ;Section 4.1
                            Status-Line        ;Section 5.1




        message-header  =  ( general-header
                           | request-header
                           | response-header
                           | entity-header )



   In the interest of robustness, any leading empty line(s) MUST be
   ignored. In other words, if the Request or Response message begins
   with one or more CRLF, CR, or LFs, these characters MUST be ignored.

4 Request

   The Request message format is shown below:



        Request  =  Request-Line       ;  Section 4.1
                    *( general-header
                    | request-header
                    | entity-header )
                    CRLF
                    [ message-body ]   ;  Section 8


4.1 Request-Line

   The Request-Line begins with a method token, followed by the
   Request-URI and the protocol version, and ending with CRLF. The
   elements are separated by SP characters.  No CR or LF are allowed
   except in the final CRLF sequence.



        Request-Line  =  Method SP Request-URI SP SIP-Version CRLF
        Request-URI   =  SIP-URL | absoluteURI


4.2 Methods




Handley/Schulzrinne/Schooler/Rosenberg                       [Page 24]

Internet Draft                    SIP                      July 13, 2000




        general-header   =  Accept               ; Section 6.7
                         |  Accept-Encoding      ; Section 6.8
                         |  Accept-Language      ; Section 6.9
                         |  Call-ID              ; Section 6.13
                         |  Contact              ; Section 6.15
                         |  CSeq                 ; Section 6.21
                         |  Date                 ; Section 6.22
                         |  Encryption           ; Section 6.23
                         |  From                 ; Section 6.25
                         |  Organization         ; Section 6.30
                         |  Record-Route         ; Section 6.35
                         |  Require              ; Section 6.36
                         |  Supported            ; Section 6.42
                         |  Timestamp            ; Section 6.43
                         |  To                   ; Section 6.44
                         |  User-Agent           ; Section 6.46
                         |  Via                  ; Section 6.47
        entity-header    =  Allow                ; Section 6.11
                         |  Content-Disposition  ; Section 6.16
                         |  Content-Encoding     ; Section 6.17
                         |  Content-Language     ; Section 6.18
                         |  Content-Length       ; Section 6.19
                         |  Content-Type         ; Section 6.20
                         |  Expires              ; Section 6.24
        request-header   =  Authorization        ; Section 6.12
                         |  Hide                 ; Section 6.26
                         |  In-Reply-To          ; Section 6.27
                         |  Max-Forwards         ; Section 6.28
                         |  Priority             ; Section 6.31
                         |  Proxy-Authorization  ; Section 6.33
                         |  Proxy-Require        ; Section 6.34
                         |  Route                ; Section 6.39
                         |  Response-Key         ; Section 6.37
                         |  Subject              ; Section 6.41
        response-header  =  Proxy-Authenticate   ; Section 6.32
                         |  Retry-After          ; Section 6.38
                         |  Server               ; Section 6.40
                         |  Unsupported          ; Section 6.45
                         |  Warning              ; Section 6.48
                         |  WWW-Authenticate     ; Section 6.49


   Table 3: SIP headers






Handley/Schulzrinne/Schooler/Rosenberg                       [Page 25]

Internet Draft                    SIP                      July 13, 2000


   The methods are defined below. Methods that are not supported by a
   proxy or redirect server are treated by that server as if they were
   an OPTIONS method and forwarded accordingly. Methods that are not
   supported by a user agent server or registrar cause a 501 (Not
   Implemented) response to be returned (Section 7). As in HTTP, the
   Method token is case-sensitive.



        Method            =  "INVITE" | "ACK" | "OPTIONS" | "BYE"
                             | "CANCEL" | "REGISTER" | extension-method
        extension-method  =  token


4.2.1 INVITE

   The INVITE method indicates that the user or service is being invited
   to participate in a session. The message body MAY contain a
   description of the session to which the callee is being invited. For
   two-party calls, the caller indicates the type of media it is able to
   receive and possibly the media it is willing to send as well as their
   parameters such as network destination. A success response MUST
   indicate in its message body which media the callee wishes to receive
   and MAY indicate the media the callee is going to send.


        Not all session description formats have the ability to
        indicate sending media.

   The caller MAY choose to omit the request body (i.e., not send a
   session description) or send a session description that does not list
   any media types. This indicates that the caller does not know its
   desired media characteristics until the call has been accepted. In
   this case, the UAS SHOULD still return a session description in its
   informational (1xx) or success (2xx) response, containing those media
   streams and codecs it supports.

   If the INVITE request did not contain a complete session description,
   the caller MUST include one in the ACK request. A UAC MUST NOT send
   an updated session description in an ACK request if it had already
   sent a complete session description in the INVITE request.  If the
   UAC wishes to modify the session after the call setup has begun, it
   MUST use another INVITE request instead.


        Delaying the session description until the ACK request is
        useful for gateways from H.323v1 to SIP, where the H.323
        media characteristics are not known until the call is



Handley/Schulzrinne/Schooler/Rosenberg                       [Page 26]

Internet Draft                    SIP                      July 13, 2000


        established.

   A server MAY automatically respond to an invitation for a conference
   the user is already participating in, identified either by the SIP
   Call-ID or a globally unique identifier within the session
   description, with a 200 (OK) response.

   The behavior of UAS depend on whether they are Internet telephony
   gateways to the PSTN. A UAS not acting as a gateway which receives an
   INVITE with a Request-URI that does not correspond to one of its
   configured addresses, MUST respond with 404 (Not Found).

   A UAS acting as a gateway translates the INVITE request into a
   telephony signaling message. If the INVITE has a Call-ID value that
   matches a recent call, the UAS compares the Request-URI with the
   Request-URI of the previous INVITE request for the same Call-ID. If
   the Request-URI contains additional digits in the "user" part, the
   UAS treats the INVITE as adding additional digits to the original
   dialed string. This is known as overlap dialing.

   If the gateway knows that the telephone number is incomplete, it
   returns a 484 (Address Incomplete) status response.

   If a user agent receives an INVITE request for an existing call leg
   with a higher CSeq sequence number than any previous INVITE for the
   same Call-ID, it MUST check any version identifiers in the session
   description or, if there are no version identifiers, the content of
   the session description to see if it has changed. It MUST also
   inspect any other header fields for changes. If there is a change,
   the user agent MUST update any internal state or information
   generated as a result of that header. If the session description has
   changed, the user agent server MUST adjust the session parameters
   accordingly, possibly after asking the user for confirmation.
   (Versioning of the session description can be used to accommodate the
   capabilities of new arrivals to a conference, add or delete media or
   change from a unicast to a multicast conference.)

   If an INVITE request for an existing session fails, the session
   description agreed upon in the last successful INVITE transaction
   remains in force.

   A UAC MUST NOT issue another INVITE request for the same call leg
   before the previous transaction has completed. A UAS that receives an
   INVITE before it sent the final response to an INVITE with a lower
   CSeq number MUST return a 400 (Bad Request) response and MUST include
   a Retry-After header field with a randomly chosen value of between 0
   and 10 seconds.




Handley/Schulzrinne/Schooler/Rosenberg                       [Page 27]

Internet Draft                    SIP                      July 13, 2000


   If a UA A sends an INVITE request to B and receives an INVITE request
   from B before it has received the response to its request from B, A
   MAY return a 500 (Internal Server Error), which SHOULD include a
   Retry-After header field specifying when the request should be
   resubmitted.


        In most cases, a UA can assume that the order of messages
        received corresponds to the order they were sent. In rare
        circumstances, the response from B and the request from B
        may be reordered on the wire.

   In addition, if A or B change multicast addresses, strict transaction
   ordering is necessary so that both sides agree on the final result.

   A UAC MUST be prepared to receive media data according to the session
   description as soon as it sends an INVITE (or re-INVITE) and can
   start sending media data when it receives a provisional or final
   response containing a session description.

   The initial INVITE from the UAC SHOULD contain the Allow and
   Supported header fields, and MAY contain the Accept header field. A
   200 (OK) response to the initial INVITE for a call SHOULD contain the
   Allow and Supported header fields, and MAY contain the Accept header
   field.


        Including these header fields allows the UAC to determine
        the features and extensions supported by the UAS for the
        duration of the call, without probing.

   This method MUST be supported by SIP proxy, redirect and user agent
   servers as well as clients.

4.2.2 ACK

   The ACK request confirms that the client has received a final
   response to an INVITE request. (ACK is used only with INVITE
   requests.) 2xx responses are acknowledged by client user agents, all
   other final responses by the first proxy or client user agent to
   receive the response. The Via is always initialized to the host that
   originates the ACK request, i.e., the client user agent after a 2xx
   response or the first proxy to receive a non-2xx final response. The
   ACK request is forwarded as the corresponding INVITE request, based
   on its Request-URI. See Section 10 for details.

   The ACK request MAY contain a message body with the final session
   description to be used by the callee. If the ACK message body is



Handley/Schulzrinne/Schooler/Rosenberg                       [Page 28]

Internet Draft                    SIP                      July 13, 2000


   empty, the callee uses the session description in the INVITE request.
   See Section 4.2.1 for further details on the relationship between
   session descriptions in INVITE and ACK requests.

   A proxy server receiving an ACK request after having sent a 3xx, 4xx,
   5xx, or 6xx response must make a determination about whether the ACK
   is for it, or for some user agent or proxy server further downstream.
   This determination is made by examining the tag in the To field. If
   the tag in the ACK To header field matches the tag in the To header
   field of the response, and the From, CSeq and Call-ID header fields
   in the response match those in the ACK, the ACK is meant for the
   proxy server. Otherwise, the ACK SHOULD be proxied downstream as any
   other request.


        It is possible for a user agent client or proxy server to
        receive multiple 3xx, 4xx, 5xx, and 6xx responses to a
        request along a single branch. This can happen under
        various error conditions, typically when a forking proxy
        transitions from stateful to stateless before receiving all
        responses. The various responses will all be identical,
        except for the tag in the To field, which is different for
        each one. It can therefore be used as a means to
        disambiguate them.

   This method MUST be supported by SIP proxy, redirect and user agent
   servers as well as clients.

4.2.3 OPTIONS

   The server is being queried as to its capabilities. A server that
   believes it can contact the user, such as a user agent where the user
   is logged in and has been recently active, MAY respond to this
   request with a capability set. A called user agent MAY return a
   status reflecting how it would have responded to an invitation, e.g.,
   600 (Busy).  A server SHOULD return Allow, Accept, Accept-Encoding,
   Accept-Language and Supported header fields. The response MAY contain
   a message body indicating the capabilities of the end system (rather
   than properties of any existing call).

   The use of the Call-ID header field is discussed in Section 6.13. An
   OPTIONS requests for an existing call-id has no impact on that call.

   Proxy and redirect servers simply forward the request without
   indicating their capabilities. This method MUST be supported by SIP
   proxy, redirect and user agent servers, registrars and clients.

4.2.4 BYE



Handley/Schulzrinne/Schooler/Rosenberg                       [Page 29]

Internet Draft                    SIP                      July 13, 2000


   The user agent client uses BYE to indicate to the server that it
   wishes to release the call. A BYE request is forwarded like an INVITE
   request and MAY be issued by either caller or callee. A party to a
   call SHOULD issue a BYE request before releasing a call ("hanging
   up"). A party receiving a BYE request MUST cease transmitting media
   streams specifically directed at the party issuing the BYE request.

   A BYE request from either called or calling party terminates any
   pending INVITE at a UA, but the INVITE request transaction MUST be
   completed with a final response and ACK.

   If the INVITE request contained a Contact header, the callee SHOULD
   send a BYE request to that address rather than the From address.

   This method MUST be supported by proxy servers and SHOULD be
   supported by redirect and user agent SIP servers.

4.2.5 CANCEL

   The CANCEL request cancels a pending request with the same Call-ID,
   To, From and CSeq (sequence number only) header field values, but
   does not affect a completed request or existing calls. (A request is
   considered completed if the server has returned a final status
   response.)

   A user agent client or proxy client MAY issue a CANCEL request at any
   time. A proxy, in particular, MAY choose to send a CANCEL to
   destinations that have not yet returned a final response after it has
   received a 2xx or 6xx response for one or more of the parallel-search
   requests. A proxy that receives a CANCEL request forwards the request
   to all destinations with pending requests.

   The Call-ID, To, the numeric part of CSeq and From header fields in
   the CANCEL request are identical to those in the original request.
   This allows a CANCEL request to be matched with the request it
   cancels. However, to allow the client to distinguish responses to the
   CANCEL from those to the original request, the CSeq Method component
   is set to CANCEL. The Via header field is initialized to the proxy
   issuing the CANCEL request. (Thus, responses to this CANCEL request
   only reach the issuing proxy.)

   The behavior of the user agent or redirect server on receiving a
   CANCEL request depends on whether the server has already sent a final
   response for the original request. If it has, the CANCEL request has
   no effect on the original request, any call state and on the
   responses generated for the original request. If the server has not
   issued a final response for the original request, it immediately
   sends a 487 (Request Terminated) for the original request. The UAC



Handley/Schulzrinne/Schooler/Rosenberg                       [Page 30]

Internet Draft                    SIP                      July 13, 2000


   confirms receipt of any final response for the original request as
   normal with an ACK request. CANCEL request itself is answered with a
   200 (OK) response in either case.


        The BYE request cannot be used to cancel branches of a
        parallel search, since several branches may, through
        intermediate proxies, find the same user agent server and
        then terminate the call.  To terminate a call instead of
        just pending searches, the UAC must use BYE instead of or
        in addition to CANCEL. While CANCEL can terminate any
        pending request other than ACK or CANCEL, it is typically
        useful only for INVITE. 200 responses to INVITE and 200
        responses to CANCEL can be distinguished by the method in
        the Cseq header field.

   This method MUST be supported by proxy servers and SHOULD be
   supported by all other SIP server types.

4.2.6 REGISTER

   A client uses the REGISTER method to register the address listed in
   the To header field with a SIP server.

   A user agent SHOULD register with a local server on startup and
   periodically thereafter by sending a REGISTER request. The period is
   given by the expiration time indicated in the registration response.
   It is RECOMMENDED that the UA registers via multicast and send a
   registration to its "home" address, i.e., the server for the domain
   that it uses as its From address in outgoing requests.

   Multicast registrations are addressed to the well-known "all SIP
   servers" multicast address "sip.mcast.net" (224.0.1.75). This request
   SHOULD be scoped to ensure it is not forwarded beyond the boundaries
   of the administrative system. This MAY be done with either TTL or
   administrative scopes [29], depending on what is implemented in the
   network. SIP user agents MAY listen to that address and use it to
   become aware of the location of other local users [20]; however, they
   do not respond to the request.

   Since registrations expire, clients need to periodically refresh
   their registrations. Such refreshes SHOULD be sent to the same
   address as the original registration, unless redirected.

   Registration requests may be redirected by a 3xx response, or with a
   Contact header in a non-200 response (typically, 401) to a REGISTER
   request. Refreshes for the same address of record SHOULD be directed
   to this new address for all subsequent registrations. A client MAY



Handley/Schulzrinne/Schooler/Rosenberg                       [Page 31]

Internet Draft                    SIP                      July 13, 2000


   revert to the original address upon reboot or upon an
   administratively configured lifetime. This implies that servers MUST
   be prepared to accept registrations on the original configured
   address and the redirected address.

   It is RECOMMENDED that clients ignore redirection responses from
   untrusted hosts or require that such redirection responses are
   cryptographically signed.


        Multicast registration may be inappropriate in some
        environments, for example, if multiple businesses share the
        same local area network.

   Requests are processed in the order received. Clients SHOULD avoid
   sending a new registration (as opposed to a retransmission) until
   they have received the response from the server for the previous one.


        Clients may register from different locations, by necessity
        using different Call-ID values. Thus, the CSeq value cannot
        be used to enforce ordering. Since registrations are
        additive, ordering is less of a problem than if each
        REGISTER request completely replaced all earlier ones.

   The meaning of the REGISTER request-header fields is defined as
   follows. We define "address-of-record" as the SIP address that the
   registry knows the registrand, typically of the form "user@domain"
   rather than "user@host". In third-party registration, the entity
   issuing the request is different from the entity being registered.

        To: The To header field contains the address-of-record whose
             registration is to be created or updated.

        From: The From header field contains the address-of-record of
             the person responsible for the registration. For first-
             party registration, it is identical to the To header field
             value.

        Request-URI: The Request-URI names the destination of the
             registration request, i.e., the domain of the registrar.
             The user name MUST be empty. Generally, the domains in the
             Request-URI and the To header field have the same value;
             however, it is possible to register as a "visitor", while
             maintaining one's name. For example, a traveler
             sip:alice@acme.com (To) might register under the Request-
             URI sip:atlanta.hiayh.org , with the former as the To
             header field and the latter as the Request-URI.  The



Handley/Schulzrinne/Schooler/Rosenberg                       [Page 32]

Internet Draft                    SIP                      July 13, 2000


             REGISTER request is no longer forwarded once it has reached
             the server whose authoritative domain is the one listed in
             the Request-URI.

        Call-ID: All registrations from a client SHOULD use the same
             Call-ID header value, at least within the same reboot
             cycle.

        Cseq: Registrations with the same Call-ID MUST have increasing
             CSeq header values. However, the server does not reject
             out-of-order requests.

        Contact: The request MAY contain a Contact header field. Future
             non-REGISTER requests for the URI given in the To header
             field SHOULD be directed to the address(es) given in the
             Contact header.

             If the request does not contain a Contact header, the
             registration remains unchanged.

             This is useful to obtain the current list of
             registrations in the response.  If a SIP URI in a
             registration Contact header field differs from
             existing registrations according to the rules in
             Section 2.1, it is added to the list of registration.
             If it is equivalent, according to these rules, to an
             existing registration, all Contact header field
             parameters for this entry are updated accordingly.
             URIs other than SIP URIs are compared according to the
             standard URI equivalency rules for the URI schema.

             All current registrations MUST share the same action value.
             Registrations that have a different action than current
             registrations for the same user MUST be rejected with
             status of 409 (Conflict).

             A proxy server ignores the q parameter when processing
             non-REGISTER requests, while a redirect server simply
             returns that parameter in its Contact response header
             field.


             Having the proxy server interpret the q parameter is
             not sufficient to guide proxy behavior, as it is not
             clear, for example, how long it is supposed to wait
             between trying addresses.

             If the registration is changed while a user agent or proxy



Handley/Schulzrinne/Schooler/Rosenberg                       [Page 33]

Internet Draft                    SIP                      July 13, 2000


             server processes an invitation, the new information SHOULD
             be used.


             This allows a service known as "directed pick-up". In
             the telephone network, directed pickup permits a user
             at a remote station who hears his own phone ringing to
             pick up at that station, dial an access code, and be
             connected to the calling user as if he had answered
             his own phone.

             Responses MAY contain a Contact header field. Its
             interpretation differs depending on the response status. If
             contained within a 2xx response, the header field indicates
             the list of current registrations. In 3xx and 4xx
             responses, it indicates where future REGISTER requests
             should be directed.

   A server MAY choose any duration for the registration lifetime.
   Registrations not refreshed after this amount of time SHOULD be
   silently discarded. Responses to a registration SHOULD include an
   Expires header (Section 6.24) or expires Contact parameters (Section
   6.15), indicating the time at which the server will drop the
   registration. If none is present, one hour is assumed. Clients MAY
   request a registration lifetime by indicating the time in an Expires
   header in the request. A server SHOULD NOT use a higher lifetime than
   the one requested, but MAY use a lower one. A single address (if
   host-independent) MAY be registered from several different clients.

   A client cancels an existing registration by sending a REGISTER
   request with an expiration time (Expires) of zero seconds, either for
   a particular Contact or the wildcard Contact designated by a "*" if
   it wants to cancel all registrations. Registrations are matched based
   on the user, host, port and maddr parameters.

   The server SHOULD return the current list of registrations in the 200
   response as Contact header fields.

   It is particularly important that REGISTER requests are authenticated
   since they allow to redirect future requests (see Section 13.2).


        Beyond its use as a simple location service, this method is
        needed if there are several SIP servers on a single host.
        In that case, only one of the servers can use the default
        port number.

   Support of this method is RECOMMENDED.



Handley/Schulzrinne/Schooler/Rosenberg                       [Page 34]

Internet Draft                    SIP                      July 13, 2000


4.3 Request-URI

   The Request-URI is a SIP URL as described in Section 2 or a general
   URI (RFC 2396 [12]).  In particular, it MUST NOT contain unescaped
   spaces or control characters. It indicates the user or service to
   which this request is being addressed. Unlike the To field, the
   Request-URI MAY be re-written by proxies.

   As shown in Table 2, the Request-URI MAY contain the user-param
   parameter as well as transport-related parameters. A server that
   receives a SIP-URL with illegal elements removes them before further
   processing.


        Transport-related parameters are needed when a UAC proxies
        all requests to a default proxy, which would then need this
        information to generate the appropriate request.


        Typically, the UAC sets the Request-URI and To to the same
        SIP URL, presumed to remain unchanged over long time
        periods. However, if the UAC has cached a more direct path
        to the callee, e.g., from the Contact header field of a
        response to a previous request, the To would still contain
        the long-term, "public" address, while the Request-URI
        would be set to the cached address.

   Proxy and redirect servers MAY use the information in the Request-URI
   and request header fields to handle the request and possibly rewrite
   the Request-URI. For example, a request addressed to the generic
   address sip:sales@acme.com is proxied to the particular person, e.g.,
   sip:bob@ny.acme.com , with the To field remaining as
   sip:sales@acme.com.  At ny.acme.com , Bob then designates Alice as
   the temporary substitute.

   The host part of the Request-URI typically agrees with one of the
   host names of the receiving server. If it does not, the server SHOULD
   proxy the request to the address indicated or return a 404 (Not
   Found) response if it is unwilling or unable to do so. For example,
   the Request-URI and server host name can disagree in the case of a
   firewall proxy that handles outgoing calls. This mode of operation is
   similar to that of HTTP proxies.

   SIP servers MAY support Request-URIs with schemes other than "sip",
   for example the "tel" URI scheme [21]. It MAY translate non-SIP URIs
   using any mechanism at its disposal, resulting in either a SIP URI or
   some other scheme.




Handley/Schulzrinne/Schooler/Rosenberg                       [Page 35]

Internet Draft                    SIP                      July 13, 2000


   If a SIP server receives a request with a URI indicating a scheme the
   server does not understand, the server MUST return a 400 (Bad
   Request) response. It MUST do this even if the To header field
   contains a scheme it does understand, since proxies are responsible
   for processing the Request-URI. (The To field is only of interest to
   the UAS.)

4.3.1 SIP Version

   Both request and response messages include the version of SIP in use,
   and follow [H3.1] (with HTTP replaced by SIP, and HTTP/1.1 replaced
   by SIP/2.0) regarding version ordering, compliance requirements, and
   upgrading of version numbers. To be compliant with this
   specification, applications sending SIP messages MUST include a SIP-
   Version of "SIP/2.0".

4.4 Option Tags

   Option tags are unique identifiers used to designate new options in
   SIP.  These tags are used in Require (Section 6.36), Supported
   (Section 6.42) and Unsupported (Section 6.45) header fields.

   Syntax:


        option-tag  =  token


   See Section C for the definition of token. The creator of a new SIP
   option MUST either prefix the option with their reverse domain name
   or register the new option with the Internet Assigned Numbers
   Authority (IANA).

   An example of a reverse-domain-name option is "com.foo.mynewfeature",
   whose inventor can be reached at "foo.com". For these features,
   individual organizations are responsible for ensuring that option
   names do not collide within the same domain. The host name part of
   the option MUST use lower-case; the option name is case-sensitive.

   Options registered with IANA do not contain periods and are globally
   unique. IANA option tags are case-sensitive.

4.4.1 Registering New Option Tags with IANA

   When registering a new SIP option, the following information MUST be
   provided:

        o Name and description of option. The name MAY be of any length,



Handley/Schulzrinne/Schooler/Rosenberg                       [Page 36]

Internet Draft                    SIP                      July 13, 2000


          but SHOULD be no more than twenty characters long. The name
          MUST consist of alphanum (See Figure 3) characters only;

        o A listing of any new SIP header fields, header parameter
          fields or parameter values defined by this option. A SIP
          option MUST NOT redefine header fields or parameters defined
          in either RFC 2543, any standards-track extensions to RFC
          2543, or other extensions registered through IANA.

        o Indication of who has change control over the option (for
          example, IETF, ISO, ITU-T, other international standardization
          bodies, a consortium or a particular company or group of
          companies);

        o A reference to a further description, if available, for
          example (in order of preference) an RFC, a published paper, a
          patent filing, a technical report, documented source code or a
          computer manual;

        o Contact information (postal and email address).

   Registrations should be sent to iana@iana.org


        This procedure has been borrowed from RTSP [4] and the RTP
        AVP [30].

5 Response

   After receiving and interpreting a request message, the recipient
   responds with a SIP response message. The response message format is
   shown below:



        Response  =  Status-Line        ;  Section 5.1
                     *( general-header
                     | response-header
                     | entity-header )
                     CRLF
                     [ message-body ]   ;  Section 8


   SIP's structure of responses is similar to [H6], but is defined
   explicitly here.

5.1 Status-Line




Handley/Schulzrinne/Schooler/Rosenberg                       [Page 37]

Internet Draft                    SIP                      July 13, 2000


   The first line of a Response message is the Status-Line, consisting
   of the protocol version (Section 4.3.1) followed by a numeric
   Status-Code and its associated textual phrase, with each element
   separated by SP characters. No CR or LF is allowed except in the
   final CRLF sequence.



        Status-Line  =  SIP-version SP Status-Code SP Reason-Phrase CRLF


5.1.1 Status Codes and Reason Phrases

   The Status-Code is a 3-digit integer result code that indicates the
   outcome of the attempt to understand and satisfy the request. The
   Reason-Phrase is intended to give a short textual description of the
   Status-Code. The Status-Code is intended for use by automata, whereas
   the Reason-Phrase is intended for the human user. The client is not
   required to examine or display the Reason-Phrase.



        Status-Code     =  Informational                     ;Fig. 4
                       |   Success                           ;Fig. 4
                       |   Redirection                       ;Fig. 5
                       |   Client-Error                      ;Fig. 6
                       |   Server-Error                      ;Fig. 7
                       |   Global-Failure                    ;Fig. 8
                       |   extension-code
        extension-code  =  3DIGIT
        Reason-Phrase   =  *


   We provide an overview of the Status-Code below, and provide full
   definitions in Section 7. The first digit of the Status-Code defines
   the class of response. The last two digits do not have any
   categorization role. SIP/2.0 allows 6 values for the first digit:

        1xx: Informational -- request received, continuing to process
             the request;

        2xx: Success -- the action was successfully received,
             understood, and accepted;

        3xx: Redirection -- further action needs to be taken in order to
             complete the request;

        4xx: Client Error -- the request contains bad syntax or cannot



Handley/Schulzrinne/Schooler/Rosenberg                       [Page 38]

Internet Draft                    SIP                      July 13, 2000


             be fulfilled at this server;

        5xx: Server Error -- the server failed to fulfill an apparently
             valid request;

        6xx: Global Failure -- the request cannot be fulfilled at any
             server.

   Figures 4 through 8 present the individual values of the numeric
   response codes, and an example set of corresponding reason phrases
   for SIP/2.0. These reason phrases are only recommended; they may be
   replaced by local equivalents without affecting the protocol. Note
   that SIP adopts many HTTP/1.1 response codes. SIP/2.0 adds response
   codes in the range starting at x80 to avoid conflicts with newly
   defined HTTP response codes, and adds a new class, 6xx, of response
   codes.

   SIP response codes are extensible. SIP applications are not required
   to understand the meaning of all registered response codes, though
   such understanding is obviously desirable. However, applications MUST
   understand the class of any response code, as indicated by the first
   digit, and treat any unrecognized response as being equivalent to the
   x00 response code of that class, with the exception that an
   unrecognized response MUST NOT be cached. For example, if a client
   receives an unrecognized response code of 431, it can safely assume
   that there was something wrong with its request and treat the
   response as if it had received a 400 (Bad Request) response code. In
   such cases, user agents SHOULD present to the user the message body
   returned with the response, since that message body is likely to
   include human-readable information which will explain the unusual
   status.



        Informational  =  "100"  ;  Trying
                      |   "180"  ;  Ringing
                      |   "181"  ;  Call Is Being Forwarded
                      |   "182"  ;  Queued
                      |   "183"  ;  Session Progress
        Success        =  "200"  ;  OK


   Figure 4: Informational and success status codes








Handley/Schulzrinne/Schooler/Rosenberg                       [Page 39]

Internet Draft                    SIP                      July 13, 2000




        Redirection  =  "300"  ;  Multiple Choices
                    |   "301"  ;  Moved Permanently
                    |   "302"  ;  Moved Temporarily
                    |   "305"  ;  Use Proxy
                    |   "380"  ;  Alternative Service


   Figure 5: Redirection status codes




        Client-Error  =  "400"  ;  Bad Request
                     |   "401"  ;  Unauthorized
                     |   "402"  ;  Payment Required
                     |   "403"  ;  Forbidden
                     |   "404"  ;  Not Found
                     |   "405"  ;  Method Not Allowed
                     |   "406"  ;  Not Acceptable
                     |   "407"  ;  Proxy Authentication Required
                     |   "408"  ;  Request Timeout
                     |   "409"  ;  Conflict
                     |   "410"  ;  Gone
                     |   "411"  ;  Length Required
                     |   "413"  ;  Request Entity Too Large
                     |   "414"  ;  Request-URI Too Large
                     |   "415"  ;  Unsupported Media Type
                     |   "420"  ;  Bad Extension
                     |   "480"  ;  Temporarily not available
                     |   "481"  ;  Call Leg/Transaction Does Not Exist
                     |   "482"  ;  Loop Detected
                     |   "483"  ;  Too Many Hops
                     |   "484"  ;  Address Incomplete
                     |   "485"  ;  Ambiguous
                     |   "486"  ;  Busy Here
                     |   "487"  ;  Request Cancelled
                     |   "488"  ;  Not Acceptable Here


   Figure 6: Client error status codes




6 Header Field Definitions




Handley/Schulzrinne/Schooler/Rosenberg                       [Page 40]

Internet Draft                    SIP                      July 13, 2000




        Server-Error  =  "500"  ;  Internal Server Error
                     |   "501"  ;  Not Implemented
                     |   "502"  ;  Bad Gateway
                     |   "503"  ;  Service Unavailable
                     |   "504"  ;  Gateway Time-out
                     |   "505"  ;  SIP Version not supported


   Figure 7: Server error status codes




        Global-Failure  =  "600"  ;  Busy Everywhere
                       |   "603"  ;  Decline
                       |   "604"  ;  Does not exist anywhere
                       |   "606"  ;  Not Acceptable


   Figure 8: Global failure status codes


   SIP header fields are similar to HTTP header fields in both syntax
   and semantics. In particular, SIP header fields follow the syntax for
   message-header as described in [H4.2]. The rules for extending header
   fields over multiple lines, and use of multiple message-header fields
   with the same field-name, described in [H4.2] also apply to SIP. The
   rules in [H4.2] regarding ordering of header fields apply to SIP,
   with the exception of Via fields, see below, whose order matters.
   Additionally, if authentication is used, the rules in Section 13.2
   apply.  Proxies SHOULD NOT reorder header fields. Proxies add Via
   header fields and MAY add other hop-by-hop header fields. They can
   modify certain header fields, such as Max-Forwards (Section 6.28) and
   "fix up" the Via header fields with "received" parameters as
   described in Section 6.47.1. Proxies MUST NOT alter any fields that
   are authenticated (see Section 13.2).

   The header fields required, optional and not applicable for each
   method are listed in Table 4 and Table 5. The table uses "o" to
   indicate optional, "m" mandatory and "-" for not applicable. A "*"
   indicates that the header fields are needed only if message body is
   not empty. See sections 6.19, 6.20 and 8 for details.

   The "where" column describes the request and response types with
   which the header field can be used. "R" refers to header fields that
   can be used in requests (that is, request and general header fields).



Handley/Schulzrinne/Schooler/Rosenberg                       [Page 41]

Internet Draft                    SIP                      July 13, 2000


   "r" designates a response or general-header field as applicable to
   all responses, while a list of numeric values indicates the status
   codes with which the header field can be used. "g" and "e" designate
   general (Section 6.1) and entity header (Section 6.2) fields,
   respectively. If a header field is marked "c", it is copied from the
   request to the response.

   The "enc." column describes whether this message header field MAY be
   encrypted end-to-end. A "n" designates fields that MUST NOT be
   encrypted, while "e" designates fields that SHOULD be encrypted if
   encryption is used.

   The "e-e" column has a value of "e" for end-to-end and a value of "h"
   for hop-by-hop header fields.



   Other header fields can be added as required; a server MUST ignore
   header fields not defined in this specification that it does not
   understand. A proxy MUST NOT remove or modify header fields not
   defined in this specification that it does not understand. A compact
   form of these header fields is also defined in Section 9 for use over
   UDP when the request has to fit into a single packet and size is an
   issue.

   Table 6 in Appendix A lists those header fields that different client
   and server types MUST be able to parse.

6.1 General Header Fields

   General header fields apply to both request and response messages.
   The "general-header" field names can be extended reliably only in
   combination with a change in the protocol version. However, new or
   experimental header fields MAY be given the semantics of general
   header fields if all parties in the communication recognize them to
   be "general-header" fields. Unrecognized header fields are treated as
   "entity-header" fields.

6.2 Entity Header Fields

   The "entity-header" fields define meta-information about the
   message-body or, if no body is present, about the resource identified
   by the request. The term "entity header" is an HTTP 1.1 term where
   the response body can contain a transformed version of the message
   body.  The original message body is referred to as the "entity". We
   retain the same terminology for header fields but usually refer to
   the "message body" rather then the entity as the two are the same in
   SIP.



Handley/Schulzrinne/Schooler/Rosenberg                       [Page 42]

Internet Draft                    SIP                      July 13, 2000



                            where  enc.  e-e ACK BYE CAN INV OPT REG
       _____________________________________________________________
       Accept                 R           e   -   o   o   o   o   o
       Accept                415          e   -   o   o   o   o   o
       Accept                 r           e   -   -   -   o   o   o
       Accept-Encoding        R           e   -   o   o   o   o   o
       Accept-Encoding       415          e   -   o   o   o   o   o
       Accept-Language        R           e   -   o   o   o   o   o
       Accept-Language       415          e   -   o   o   o   o   o
       Alert-Info             R     e     e   -   -   -   o   -   -
       Allow                 200          e   -   -   -   o   o   o
       Allow                 405          e   m   m   m   m   m   m
       Authorization          R           e   o   o   o   o   o   o
       Authorization          r           e   o   o   o   o   o   o
       Call-ID               gc     n     e   m   m   m   m   m   m
       Call-Info              g     e     e   -   -   -   o   o   o
       Contact                R           e   o   -   -   o   o   o
       Contact               1xx          e   -   -   -   o   o   -
       Contact               2xx          e   -   -   -   o   o   o
       Contact               3xx          e   -   o   -   o   o   o
       Contact               485          e   -   o   -   o   o   o
       Content-Disposition    e           e   o   o   -   o   o   o
       Content-Encoding       e           e   o   o   -   o   o   o
       Content-Language       e           e   m   m   m   m   m   m
       Content-Length         e     n     e   m   m   m   m   m   m
       Content-Type           e           e   *   *   -   *   *   *
       CSeq                  gc     n     e   m   m   m   m   m   m
       Date                   g           e   o   o   o   o   o   o
       Encryption             g     n     e   o   o   o   o   o   o
       Expires                g           e   -   -   -   o   -   o
       From                  gc     n     e   m   m   m   m   m   m
       Hide                   R     n     h   o   o   o   o   o   o
       In-Reply-To            R     e     e   -   -   -   o   -   -
       Max-Forwards           R     n     e   o   o   o   o   o   o
       MIME-Version           g     n     e   o   o   o   o   o   o
       Organization           g     e     h   -   -   -   o   o   o


   Table 4: Summary of header fields, A--O

6.3 Request Header Fields

   The "request-header" fields allow the client to pass additional
   information about the request, and about the client itself, to the
   server. These fields act as request modifiers, with semantics
   equivalent to the parameters of a programming language method
   invocation.



Handley/Schulzrinne/Schooler/Rosenberg                       [Page 43]

Internet Draft                    SIP                      July 13, 2000



                             where       enc.  e-e ACK BYE CAN INV OPT REG
   _______________________________________________________________________
   Priority                    R          c     e   -   -   -   o   -   -
   Proxy-Authenticate       401,407       n     h   o   o   o   o   o   o
   Proxy-Authorization         R          n     h   o   o   o   o   o   o
   Proxy-Require               R          n     h   o   o   o   o   o   o
   Record-Route                R                h   o   o   o   o   o   o
   Record-Route           2xx,401,484           h   o   o   o   o   o   o
   Require                     g                e   o   o   o   o   o   o
   Response-Key                R          c     e   -   o   o   o   o   o
   Retry-After                 R          c     e   -   -   -   -   -   o
   Retry-After          404,413,480,486   c     e   o   o   o   o   o   o
                            500,503       c     e   o   o   o   o   o   o
                            600,603       c     e   o   o   o   o   o   o
   Route                       R                h   o   o   o   o   o   o
   Server                      r          c     e   o   o   o   o   o   o
   Subject                     R          c     e   -   -   -   o   -   -
   Supported                   g          c     e   -   o   o   o   o   o
   Timestamp                   g                e   o   o   o   o   o   o
   To                        gc(1)        n     e   m   m   m   m   m   m
   Unsupported                 R                e   o   o   o   o   o   o
   Unsupported                420               e   o   o   o   o   o   o
   User-Agent                  g          c     e   o   o   o   o   o   o
   Via                       gc(2)        n     e   m   m   m   m   m   m
   Warning                     r                e   o   o   o   o   o   o
   WWW-Authenticate            R          c     e   o   o   o   o   o   o
   WWW-Authenticate           401         c     e   o   o   o   o   o   o


   Table 5: Summary of header fields, P--Z; (1):  copied  with  possible
   addition of tag; (2): UAS removes first Via header field

   The "request-header" field names can be extended reliably only in
   combination with a change in the protocol version. However, new or
   experimental header fields MAY be given the semantics of "request-
   header" fields if all parties in the communication recognize them to
   be request-header fields. Unrecognized header fields are treated as
   "entity-header" fields.

6.4 Response Header Fields

   The "response-header" fields allow the server to pass additional
   information about the response which cannot be placed in the Status-
   Line. These header fields give information about the server and about
   further access to the resource identified by the Request-URI.

   Response-header field names can be extended reliably only in
   combination with a change in the protocol version. However, new or


Handley/Schulzrinne/Schooler/Rosenberg                       [Page 44]

Internet Draft                    SIP                      July 13, 2000


   experimental header fields MAY be given the semantics of "response-
   header" fields if all parties in the communication recognize them to
   be "response-header" fields. Unrecognized header fields are treated
   as "entity-header" fields.

6.5 End-to-end and Hop-by-hop Headers

   End-to-end headers MUST be transmitted unmodified across all proxies,
   while hop-by-hop headers MAY be modified or added by proxies.

6.6 Header Field Format

   Header fields ("general-header", "request-header", "response-header",
   and "entity-header") follow the same generic header format as that
   given in Section 3.1 of RFC 822 [28]. Each header field consists of a
   name followed by a colon (":") and the field value. Field names are
   case-insensitive. The field value MAY be preceded by any amount of
   leading white space (LWS), though a single space (SP) is preferred.
   Header fields can be extended over multiple lines by preceding each
   extra line with at least one SP or horizontal tab (HT). Applications
   MUST follow HTTP "common form" when generating these constructs,
   since there might exist some implementations that fail to accept
   anything beyond the common forms.



        message-header  =  field-name ":" [ field-value ] CRLF
        field-name      =  token
        field-value     =  *( field-content | LWS )
        field-content   =  < the OCTETs  making up the field-value
                            and consisting of either *TEXT-UTF8
                            or combinations of token,
                            separators, and quoted-string>


   The relative order of header fields with different field names is not
   significant. Multiple header fields with the same field-name may be
   present in a message if and only if the entire field-value for that
   header field is defined as a comma-separated list (i.e., #(values)).
   It MUST be possible to combine the multiple header fields into one
   "field-name: field-value" pair, without changing the semantics of the
   message, by appending each subsequent field-value to the first, each
   separated by a comma. The order in which header fields with the same
   field-name are received is therefore significant to the
   interpretation of the combined field value, and thus a proxy MUST NOT
   change the order of these field values when a message is forwarded.

   Unless otherwise noted, comparisons are case-sensitive. However, any



Handley/Schulzrinne/Schooler/Rosenberg                       [Page 45]

Internet Draft                    SIP                      July 13, 2000


   parameters preceded by semicolons are case-insensitive, while any
   quoted string literals are compared taken case into consideration.

   The Contact, From and To header fields contain a URL. If the URL
   contains a comma, question mark or semicolon, the URL MUST be
   enclosed in angle brackets (< and >). Any URL parameters are
   contained within these brackets. If the URL is not enclosed in angle
   brackets, any semicolon-delimited parameters are header-parameters,
   not URL parameters.

6.7 Accept

   The Accept header follows the syntax defined in [H14.1]. The
   semantics are also identical, with the exception that if no Accept
   header is present, the server SHOULD assume a default value of
   application/sdp

   As a request-header field, it is used only with those methods that
   take message bodies. In a 415 (Unsupported Media Type) response, it
   indicates which content types are acceptable in requests. In 200 (OK)
   responses for INVITE, it lists the content types acceptable for
   future requests in this call.

   Example:


     Accept: application/sdp;level=1, application/x-private, text/html



6.8 Accept-Encoding

   The Accept-Encoding request-header field is similar to Accept, but
   restricts the content-codings [H3.4.1] that are acceptable in the
   response. See [H14.3]. The syntax of this header is defined in
   [H14.3]. The semantics in SIP are identical to those defined in
   [H14.3].


        Note: An empty Accept-Encoding header field is permissible,
        even though the syntax in [H14.3] does not provide for it.
        It is equivalent to Accept-Encoding: identity, i.e., only
        the identity encoding, meaning no encoding, is permissible.

   If no Accept-Encoding header field is present in a request, the
   server MUST use the "identity" encoding.





Handley/Schulzrinne/Schooler/Rosenberg                       [Page 46]

Internet Draft                    SIP                      July 13, 2000


        HTTP/1.1 [H14.3] states that the server SHOULD use the
        "identity" encoding unless it has additional information
        about the capabilities of the client. This is needed for
        backwards-compatibility with old HTTP clients and does not
        affect SIP.

6.9 Accept-Language

   The Accept-Language header follows the syntax defined in [H14.4]. The
   rules for ordering the languages based on the q parameter apply to
   SIP as well. When used in SIP, the Accept-Language request-header
   field can be used to allow the client to indicate to the server in
   which language it would prefer to receive reason phrases, session
   descriptions or status responses carried as message bodies. A proxy
   MAY use this field to help select the destination for the call, for
   example, a human operator conversant in a language spoken by the
   caller.

   Example:


     Accept-Language: da, en-gb;q=0.8, en;q=0.7



6.10 Alert-Info

   The Alert-Info header field indicates that the content indicated in
   the URLs should be rendered instead of ring tone. A user SHOULD be
   able to disable this feature selectively to prevent unauthorized
   disruptions.



        Alert-Info  =  "Alert-Info" ":" # ( "<" URI ">" *(";" generic-param ))


   Example:

   Alert-Info: 



6.11 Allow

   The Allow entity-header field lists the set of methods supported by
   the resource identified by the Request-URI. The purpose of this field
   is strictly to inform the recipient of valid methods associated with



Handley/Schulzrinne/Schooler/Rosenberg                       [Page 47]

Internet Draft                    SIP                      July 13, 2000


   the resource. An Allow header field MUST be present in a 405 (Method
   Not Allowed) response, SHOULD be present in an OPTIONS response
   SHOULD be present in the 200 (OK) response to the initial INVITE for
   a call and MAY be present in final responses for other methods.


        Supplying an Allow header in responses to methods other
        than OPTIONS cuts down on the number of messages needed.



        Allow  =  "Allow" ":" 1#Method


6.12 Authorization

   A user agent that wishes to authenticate itself with a UAS or
   registrar -- usually, but not necessarily, after receiving a 401
   response -- MAY do so by including an Authorization request-header
   field with the request. The Authorization field value consists of
   credentials containing the authentication information of the user
   agent for the realm of the resource being requested.

   Section 13.2 overviews the use of the Authorization header, and
   section 15 describes the syntax and semantics when used with PGP
   based authentication.

6.13 Call-ID

   The Call-ID general-header field uniquely identifies a particular
   invitation or all registrations of a particular client. Note that a
   single multimedia conference can give rise to several calls with
   different Call-IDs, e.g., if a user invites a single individual
   several times to the same (long-running) conference.

   For an INVITE request, a callee user agent server SHOULD NOT alert
   the user if the user has responded previously to the Call-ID in the
   INVITE request. If the user is already a member of the conference and
   the conference parameters contained in the session description have
   not changed, a callee user agent server MAY silently accept the call,
   regardless of the Call-ID. An invitation for an existing Call-ID or
   session can change the parameters of the conference. A client
   application MAY decide to simply indicate to the user that the
   conference parameters have been changed and accept the invitation
   automatically or it MAY require user confirmation.

   A user may be invited to the same conference or call using several
   different Call-IDs. If desired, the client MAY use identifiers within



Handley/Schulzrinne/Schooler/Rosenberg                       [Page 48]

Internet Draft                    SIP                      July 13, 2000


   the session description to detect this duplication. For example, SDP
   contains a session id and version number in the origin (o) field.

   The REGISTER and OPTIONS methods use the Call-ID value (in addition
   to the CSeq value) to unambiguously match requests and responses. All
   REGISTER requests issued by a single client SHOULD use the same
   Call-ID, at least within the same boot cycle. For these requests, it
   makes no difference whether the Call-ID value matches an existing
   call or not.


        Since the Call-ID is generated by and for SIP, there is no
        reason to deal with the complexity of URL-encoding and
        case-ignoring string comparison.



        callid   =  token [ "@" token ]
        Call-ID  =  ( "Call-ID" | "i" ) ":" callid


   The callid MUST be a globally unique identifier and MUST NOT be
   reused for later calls. Use of cryptographically random identifiers
   [31] is RECOMMENDED. Implementations MAY use the form "localid@host".
   Call-IDs are case-sensitive and are simply compared byte-by-byte.


        Using cryptographically random identifiers provides some
        protection against session hijacking. Call-ID, To and From
        are needed to identify a call leg.  The distinction between
        call and call leg matters in calls with third-party
        control.

   For systems which have tight bandwidth constraints, many of the
   mandatory SIP headers have a compact form, as discussed in Section 9.
   These are alternate names for the headers which occupy less space in
   the message. In the case of Call-ID, the compact form is i.

   For example, both of the following are valid:

     Call-ID: f81d4fae-7dec-11d0-a765-00a0c91e6bf6@foo.bar.com


   or

     i:f81d4fae-7dec-11d0-a765-00a0c91e6bf6@foo.bar.com





Handley/Schulzrinne/Schooler/Rosenberg                       [Page 49]

Internet Draft                    SIP                      July 13, 2000


6.14 Call-Info

   The Call-Info general header field provides additional information
   about the caller or callee, depending on whether it is found in a
   request or response. The purpose of the URI is described by the
   "purpose" parameter. "icon" designates an image suitable as an iconic
   representation of the caller or callee; "info" describes the caller
   or callee in general, e.g., through a web page; "card" provides a
   business card (e.g., in vCard [32] or LDIF [33] formats).



        Call-Info  =  "Call-Info" ":" # ( "<" URI ">"
                      [ ";" purpose "=" ( "icon" | "info" | "card" | token) ]
                      *(";" generic-param ))


   Example:

   Call-Info:  ;purpose=icon,
      ;purpose=info



6.15 Contact

   The Contact general-header field can appear in INVITE, OPTIONS, ACK,
   and REGISTER requests, and in 1xx, 2xx, 3xx, and 485 responses. In
   general, it provides a URL where the user can be reached for further
   communications.

   In some of the cases below, the client uses information from the
   Contact header field in Request-URI of future requests. In these
   cases, the client copies all but the "method-param" and "header"
   elements of the addr-spec part of the Contact header field into the
   Request-URI of the request. It uses the "header" parameters to create
   headers for the request, replacing any default headers normally used.
   Unless the client is configured to use a default proxy for all
   outgoing requests, it then directs the request to the address and
   port specified by the "maddr" and "port" parameters, using the
   transport protocol given in the "transport" parameter. If "maddr" is
   a multicast address, the value of "ttl" is used as the time-to-live
   value.

        INVITE, OPTIONS and ACK requests: INVITE requests MUST and ACK
             requests MAY contain Contact headers indicating from which
             location the request is originating. The URL in the Contact
             header field is then used by subsequent requests from the



Handley/Schulzrinne/Schooler/Rosenberg                       [Page 50]

Internet Draft                    SIP                      July 13, 2000


             callee. For OPTIONS, Contact provides a hint where future
             SIP requests can be sent or the user can be contacted via
             non-SIP means.


             This allows the callee to send future requests, such
             as BYE, directly to the caller instead of through a
             series of proxies.  The Via header is not sufficient
             since the desired address may be that of a proxy.

        INVITE 1xx responses: A UAS sending a provisional response (1xx)
             MAY insert a Contact response header. It has the same
             semantics in a 1xx response as a 2xx INVITE response. Note
             that CANCEL requests MUST NOT be sent to that address, but
             rather follow the same path as the original request.

        INVITE and OPTIONS 2xx responses: A user agent server sending a
             definitive, positive response (2xx) MUST insert a Contact
             response header field indicating the SIP address under
             which it is reachable most directly for future SIP
             requests, such as ACK, within the same Call-ID. The Contact
             header field contains the address of the server itself or
             that of a proxy, e.g., if the host is behind a firewall.
             The value of this Contact header is copied into the
             Request-URI of subsequent requests for this call if the
             response did not also contain a Record-Route header. If the
             response also contains a Record-Route header field, the
             address in the Contact header field is added as the last
             item in the Route header field. See Section 6.35 for
             details.

             If a UA supports both UDP and TCP, it SHOULD NOT indicate a
             transport parameter in the URI.


             The Contact value SHOULD NOT be cached across calls,
             as it may not represent the most desirable location
             for a particular destination address.

        REGISTER requests: REGISTER requests MAY contain a Contact
             header field indicating at which locations the user is
             reachable. The REGISTER request defines a wildcard Contact
             field, "*", which MUST only be used with Expires: 0 to
             remove all registrations for a particular user. An optional
             "expires" parameter indicates the desired expiration time
             of the registration. If a Contact entry does not have an
             "expires" parameter, the Expires header field is used as
             the default value. If neither of these mechanisms is used,



Handley/Schulzrinne/Schooler/Rosenberg                       [Page 51]

Internet Draft                    SIP                      July 13, 2000


             SIP URIs are assumed to expire after one hour. Other URI
             schemes have no expiration times.

        REGISTER 2xx responses: A REGISTER response MAY return all
             locations at which the user is currently reachable. An
             optional "expires" parameter indicates the expiration time
             of the registration. If a Contact entry does not have an
             "expires" parameter, the value of the Expires header field
             indicates the expiration time. If neither mechanism is
             used, the expiration time specified in the request,
             explicitly or by default, is used.

        3xx and 485 responses: The Contact response-header field can be
             used with a 3xx or 485 (Ambiguous) response codes to
             indicate one or more alternate addresses to try. It can
             appear in responses to BYE, INVITE and OPTIONS methods. The
             Contact header field contains URIs giving the new locations
             or user names to try, or may simply specify additional
             transport parameters. A 300 (Multiple Choices), 301 (Moved
             Permanently), 302 (Moved Temporarily) or 485 (Ambiguous)
             response SHOULD contain a Contact field containing URIs of
             new addresses to be tried. A 301 or 302 response may also
             give the same location and username that was being tried
             but specify additional transport parameters such as a
             different server or multicast address to try or a change of
             SIP transport from UDP to TCP or vice versa.  The client
             copies information from the Contact header field into the
             Request-URI as described above.

        4xx, 5xx and 6xx responses: The Contact response-header field
             can be used with a 4xx, 5xx or 6xx response to indicate the
             location where additional information about the error can
             be found.

   Note that the Contact header field MAY also refer to a different
   entity than the one originally called. For example, a SIP call
   connected to GSTN gateway may need to deliver a special information
   announcement such as "The number you have dialed has been changed."

   A Contact response header field can contain any suitable URI
   indicating where the called party can be reached, not limited to SIP
   URLs. For example, it could contain URL's for phones, fax, or irc (if
   they were defined) or a mailto: (RFC 2368, [34]) URL.

   The following parameters are defined. Additional parameters may be
   defined in other specifications.

        q: The "qvalue" indicates the relative preference among the



Handley/Schulzrinne/Schooler/Rosenberg                       [Page 52]

Internet Draft                    SIP                      July 13, 2000


             locations given. "qvalue" values are decimal numbers from 0
             to 1, with higher values indicating higher preference.

        action: The "action" parameter is used only when registering
             with the REGISTER request. It indicates whether the client
             wishes that the server proxy or redirect future requests
             intended for the client. If this parameter is not specified
             the action taken depends on server configuration. In its
             response, the registrar SHOULD indicate the mode used. This
             parameter is ignored for other requests.

        expires: The "expires" parameter indicates how long the URI is
             valid. The parameter is either a number indicating seconds
             or a quoted string containing a SIP-date. If this parameter
             is not provided, the value of the Expires header field
             determines how long the URI is valid. Implementations MAY
             treat values larger than 2**32-1 (4294967295 seconds or 136
             years) as equivalent to 2**32-1.



   Contact = ( "Contact" | "m" ) ":" 
             ("*" | (1# (( name-addr | addr-spec )
             *( ";" contact-params ) )))

   name-addr      = [ display-name ] "<" addr-spec ">"
   addr-spec      = SIP-URL | URI
   display-name   = *token | quoted-string

   contact-params = "q"       "=" qvalue
                  | "action"  "=" "proxy" | "redirect"
                  | "expires" "=" delta-seconds | <"> SIP-date <">
                  | contact-extension

   contact-extension = generic-param


   Even if the "display-name" is empty, the "name-addr" form MUST be
   used if the "addr-spec" contains a comma, semicolon or question mark.
   Note that there may or may not be LWS between the display-name and
   the "<".


        The Contact header field fulfills functionality similar to
        the Location header field in HTTP. However, the HTTP header
        only allows one address, unquoted. Since URIs can contain
        commas and semicolons as reserved characters, they can be
        mistaken for header or parameter delimiters, respectively.



Handley/Schulzrinne/Schooler/Rosenberg                       [Page 53]

Internet Draft                    SIP                      July 13, 2000


        The current syntax corresponds to that for the To and From
        header, which also allows the use of display names.

   Example:


     Contact: "Mr. Watson" 
        ;q=0.7; expires=3600,
        "Mr. Watson"  ;q=0.1



6.16 Content-Disposition



        Content-Disposition  =  "Content-Disposition" ":"
                                disposition-type *(";" disposition-parm)
        disposition-type     =  "render" | "session" | extension-token
        disposition-parm     =  handling-parm | parameter
        handling-parm        =  "handling" "="
                                ( "optional" | "required" | other-handling )
        other-handling       =  token


   The Content-Disposition header field describes how the message body
   or, in the case of multipart messages, a message body part is to be
   interpreted by the UAC or UAS. The SIP header extends the MIME
   Content-Type (RFC 1806 [35]).

   The value "session" indicates that the body part describes a session.
   The value "render" indicates that the body part should be displayed
   or otherwise rendered to the user. For backward-compatibility,
   message bodies in requests and 2xx responses are assumed to have the
   value "session", while those in other responses have the value
   "render".

   The handling parameter, handling-parm, describes how the UAS should
   react if it receives a message body whose content type or disposition
   type it does not understand. If the parameter has the value
   "optional", the UAS MUST ignore the message body; if it has the value
   "required", the UAS MUST return 415 (Unsupported Media Type).

6.17 Content-Encoding



        Content-Encoding  =  ( "Content-Encoding" | "e" ) ":"



Handley/Schulzrinne/Schooler/Rosenberg                       [Page 54]

Internet Draft                    SIP                      July 13, 2000


                             1#content-coding


   The Content-Encoding entity-header field is used as a modifier to the
   "media-type". When present, its value indicates what additional
   content codings have been applied to the entity-body, and thus what
   decoding mechanisms MUST be applied in order to obtain the media-type
   referenced by the Content-Type header field.  Content-Encoding is
   primarily used to allow a body to be compressed without losing the
   identity of its underlying media type.

   If multiple encodings have been applied to an entity, the content
   codings MUST be listed in the order in which they were applied.

   All content-coding values are case-insensitive. The Internet Assigned
   Numbers Authority (IANA) acts as a registry for content-coding value
   tokens. See [3.5] for a definition of the syntax for content-coding.

   Clients MAY apply content encodings to the body in requests. If the
   server is not capable of decoding the body, or does not recognize any
   of the content-coding values, it MUST send a 415 "Unsupported Media
   Type" response, listing acceptable encodings in the Accept-Encoding
   header. A server MAY apply content encodings to the bodies in
   responses. The server MUST only use encodings listed in the Accept-
   Encoding header in the request.

6.18 Content-Language

   See [H14.12].

6.19 Content-Length

   The Content-Length entity-header field indicates the size of the
   message-body, in decimal number of octets, sent to the recipient.



        Content-Length  =  ( "Content-Length" | "l" ) ":" 1*DIGIT


   An example is

     Content-Length: 3495



   Applications SHOULD use this field to indicate the size of the
   message-body to be transferred, regardless of the media type of the



Handley/Schulzrinne/Schooler/Rosenberg                       [Page 55]

Internet Draft                    SIP                      July 13, 2000


   entity. Any Content-Length greater than or equal to zero is a valid
   value. If no body is present in a message, then the Content-Length
   header field MUST be set to zero. If a server receives a UDP request
   without Content-Length, it MUST assume that the request encompasses
   the remainder of the packet.  If a server receives a UDP request with
   a Content-Length, but the value is larger than the size of the body
   sent in the request, the client SHOULD generate a 400 class response.
   If there is additional data in the UDP packet after the last byte of
   the body has been read, the server MUST treat the remaining data as a
   separate message. This allows several messages to be placed in a
   single UDP packet.

   If a response does not contain a Content-Length, the client assumes
   that it encompasses the remainder of the UDP packet or the data until
   the TCP connection is closed, as applicable.  Section 8 describes how
   to determine the length of the message body.


        The ability to omit Content-Length simplifies the creation
        of cgi-like scripts that dynamically generate responses.

6.20 Content-Type

   The Content-Type entity-header field indicates the media type of the
   message-body sent to the recipient. The "media-type" element is
   defined in [H3.7].



        Content-Type  =  ( "Content-Type" | "c" ) ":" media-type


   Examples of this header field are

     Content-Type: application/sdp
     Content-Type: text/html; charset=ISO-8859-4



6.21 CSeq

   Clients MUST add the CSeq (command sequence) general-header field to
   every request. A CSeq header field in a request contains the request
   method and a single decimal sequence number chosen by the requesting
   client, unique within a single call leg.  The sequence number MUST be
   expressible as a 32-bit unsigned integer. The initial value of the
   sequence number is arbitrary, but MUST be less than 2**31.
   Consecutive requests that differ in request method, headers or body,



Handley/Schulzrinne/Schooler/Rosenberg                       [Page 56]

Internet Draft                    SIP                      July 13, 2000


   but have the same Call-ID MUST contain strictly monotonically
   increasing and contiguous sequence numbers; sequence numbers do not
   wrap around.  Retransmissions of the same request carry the same
   sequence number, but an INVITE with a different message body or
   different header fields (a "re-invitation") acquires a new, higher
   sequence number. A server MUST echo the CSeq value from the request
   in its response.  If the Method value is missing in the received CSeq
   header field, the server fills it in appropriately.

   The ACK and CANCEL requests MUST contain the same CSeq value as the
   INVITE request that it refers to, while a BYE request cancelling an
   invitation MUST have a higher sequence number. A BYE request with a
   CSeq that is not higher should cause a 400 response to be generated.

   A user agent server MUST remember the highest sequence number for any
   INVITE request with the same Call-ID value. The server MUST respond
   to, and then discard, any INVITE request with a lower sequence
   number.

   All requests spawned in a parallel search have the same CSeq value as
   the request triggering the parallel search.



        CSeq  =  "CSeq" ":" 1*DIGIT Method



        Strictly speaking, CSeq header fields are needed for any
        SIP request that can be cancelled by a BYE or CANCEL
        request or where a client can issue several requests for
        the same Call-ID in close succession. Without a sequence
        number, the response to an INVITE could be mistaken for the
        response to the cancellation (BYE or CANCEL). Also, if the
        network duplicates packets or if an ACK is delayed until
        the server has sent an additional response, the client
        could interpret an old response as the response to a re-
        invitation issued shortly thereafter. Using CSeq also makes
        it easy for the server to distinguish different versions of
        an invitation, without comparing the message body.

   The Method value allows the client to distinguish the response to an
   INVITE request from that of a CANCEL response. CANCEL requests can be
   generated by proxies; if they were to increase the sequence number,
   it might conflict with a later request issued by the user agent for
   the same call.

   With a length of 32 bits, a server could generate, within a single



Handley/Schulzrinne/Schooler/Rosenberg                       [Page 57]

Internet Draft                    SIP                      July 13, 2000


   call, one request a second for about 136 years before needing to wrap
   around.  The initial value of the sequence number is chosen so that
   subsequent requests within the same call will not wrap around. A
   non-zero initial value allows to use a time-based initial sequence
   number, if the client desires. A client could, for example, choose
   the 31 most significant bits of a 32-bit second clock as an initial
   sequence number.

   Forked requests MUST have the same CSeq as there would be ambiguity
   otherwise between these forked requests and later BYE issued by the
   client user agent.

   Example:


     CSeq: 4711 INVITE



6.22 Date

   Date is a general-header field. Its syntax is:



        Date      =  "Date" ":" SIP-date
        SIP-date  =  rfc1123-date


   See [H14.18] for a definition of rfc1123-date. Note that unlike
   HTTP/1.1, SIP only supports the most recent RFC 1123 [36] formatting
   for dates.  As in [H3.3], SIP restricts the timezone in SIP-date to
   "GMT", while RFC 1123 allows any timezone.

        The consistent use of GMT between Date, Expires and Retry-
        After headers allows implementation of simple clients that
        do not have a notion of absolute time.  Note that rfc1123-
        date is case-sensitive.

   The Date header field reflects the time when the request or response
   is first sent. Thus, retransmissions have the same Date header field
   value as the original.


        The Date header field can be used by simple end systems
        without a battery-backed clock to acquire a notion of
        current time. However, in its GMT-form, it requires clients
        to know their offset from GMT.



Handley/Schulzrinne/Schooler/Rosenberg                       [Page 58]

Internet Draft                    SIP                      July 13, 2000


6.23 Encryption

   The Encryption general-header field specifies that the content has
   been encrypted. Section 13 describes the overall SIP security
   architecture and algorithms. This header field is intended for end-
   to-end encryption of requests and responses. Requests are encrypted
   based on the public key belonging to the entity named in the To
   header field. Responses are encrypted based on the public key
   conveyed in the Response-Key header field. Note that the public keys
   themselves may not be used for the encryption. This depends on the
   particular algorithms used.

   For any encrypted message, at least the message body and possibly
   other message header fields are encrypted. An application receiving a
   request or response containing an Encryption header field decrypts
   the body and then concatenates the plaintext to the request line and
   headers of the original message. Message headers in the decrypted
   part completely replace those with the same field name in the
   plaintext part.  (Note: If only the body of the message is to be
   encrypted, the body has to be prefixed with CRLF to allow proper
   concatenation.) Note that the request method and Request-URI cannot
   be encrypted.


        Encryption only provides privacy; the recipient has no
        guarantee that the request or response came from the party
        listed in the From message header, only that the sender
        used the recipient's public key. However, proxies will not
        be able to modify the request or response.



        Encryption         =  "Encryption" ":" encryption-scheme 1*SP
                              #encryption-params
        encryption-scheme  =  token
        encryption-params  =  generic-param
        generic-param      =  token [ "=" ( token | quoted-string ) ]

        The token indicates the form of encryption used; it is
        described in section 13.

   The example in Figure 9 shows a message encrypted with ASCII-armored
   PGP that was generated by applying "pgp -ea" to the payload to be
   encrypted.



   Figure 9: PGP Encryption Example



Handley/Schulzrinne/Schooler/Rosenberg                       [Page 59]

Internet Draft                    SIP                      July 13, 2000



   INVITE sip:watson@boston.bell-telephone.com SIP/2.0
   Via: SIP/2.0/UDP 169.130.12.5
   From: 
   To: T. A. Watson 
   Call-ID: 187602141351@worcester.bell-telephone.com
   Cseq: 1 INVITE
   Content-Length: 829
   Encryption: PGP version=2.6.2,encoding=ascii

   hQEMAxkp5GPd+j5xAQf/ZDIfGD/PDOM1wayvwdQAKgGgjmZWe+MTy9NEX8O25Red
   h0/pyrd/+DV5C2BYs7yzSOSXaj1C/tTK/4do6rtjhP8QA3vbDdVdaFciwEVAcuXs
   ODxlNAVqyDi1RqFC28BJIvQ5KfEkPuACKTK7WlRSBc7vNPEA3nyqZGBTwhxRSbIR
   RuFEsHSVojdCam4htcqxGnFwD9sksqs6LIyCFaiTAhWtwcCaN437G7mUYzy2KLcA
   zPVGq1VQg83b99zPzIxRdlZ+K7+bAnu8Rtu+ohOCMLV3TPXbyp+err1YiThCZHIu
   X9dOVj3CMjCP66RSHa/ea0wYTRRNYA/G+kdP8DSUcqYAAAE/hZPX6nFIqk7AVnf6
   IpWHUPTelNUJpzUp5Ou+q/5P7ZAsn+cSAuF2YWtVjCf+SQmBR13p2EYYWHoxlA2/
   GgKADYe4M3JSwOtqwU8zUJF3FIfk7vsxmSqtUQrRQaiIhqNyG7KxJt4YjWnEjF5E
   WUIPhvyGFMJaeQXIyGRYZAYvKKklyAJcm29zLACxU5alX4M25lHQd9FR9Zmq6Jed
   wbWvia6cAIfsvlZ9JGocmQYF7pcuz5pnczqP+/yvRqFJtDGD/v3s++G2R+ViVYJO
   z/lxGUZaM4IWBCf+4DUjNanZM0oxAE28NjaIZ0rrldDQmO8V9FtPKdHxkqA5iJP+
   6vGOFti1Ak4kmEz0vM/Nsv7kkubTFhRl05OiJIGr9S1UhenlZv9l6RuXsOY/EwH2
   z8X9N4MhMyXEVuC9rt8/AUhmVQ==
   =bOW+

   Since proxies can base their forwarding decision on any combination
   of SIP header fields, there is no guarantee that an encrypted request
   "hiding" header fields will reach the same destination as an
   otherwise identical un-encrypted request.

6.24 Expires

   The Expires entity-header field gives the date and time after which
   the message content expires.

   This header field is currently defined only for the REGISTER and
   INVITE methods. For REGISTER, it is a request and response-header
   field. In a REGISTER request, the client indicates how long it wishes
   the registration to be valid. In the response, the server indicates
   the earliest expiration time of all registrations. The server MAY
   choose a shorter time interval than that requested by the client, but
   SHOULD NOT choose a longer one.  If a registration updates an
   existing registration, the Expires value of the most recent
   registration is used, even if it is shorter than the earlier
   registration.

   For INVITE requests, it is a request and response-header field. In a
   request, the caller can limit the validity of an invitation, for



Handley/Schulzrinne/Schooler/Rosenberg                       [Page 60]

Internet Draft                    SIP                      July 13, 2000


   example, if a client wants to limit the time duration of a search or
   a conference invitation. A user interface MAY take this as a hint to
   leave the invitation window on the screen even if the user is not
   currently at the workstation. This also limits the duration of a
   search. If the request expires before the search completes, the proxy
   returns a 408 (Request Timeout) status. In a 302 (Moved Temporarily)
   response, a server can advise the client of the maximal duration of
   the redirection.

   Note that the expiration time does not affect the duration of the
   actual session that may result from the invitation. Session
   description protocols may offer the ability to express time limits on
   the session duration, however.

   The value of this field can be either a SIP-date or an integer number
   of seconds (in decimal), measured from the receipt of the request.
   The latter approach is preferable for short durations, as it does not
   depend on clients and servers sharing a synchronized clock.
   Implementations MAY treat values larger than 2**32-1 (4294967295 or
   136 years) as equivalent to 2**32-1.



        Expires  =  "Expires" ":" ( SIP-date | delta-seconds )


   Two examples of its use are

     Expires: Thu, 01 Dec 1994 16:00:00 GMT
     Expires: 5



6.25 From

   Requests and responses MUST contain a From general-header field,
   indicating the initiator of the request.  (Note that this may be
   different from the initiator of the call leg. Requests sent by the
   callee to the caller use the callee's address in the From header
   field.)  The From field MAY contain the "tag" parameter.  The server
   copies the From header field from the request to the response. The
   optional "display-name" is meant to be rendered by a human-user
   interface. A system SHOULD use the display name "Anonymous" if the
   identity of the client is to remain hidden.

   The SIP-URL MUST NOT contain the "transport-param", "maddr-param",
   "ttl-param", or "headers" elements. A server that receives a SIP-URL
   with these elements ignores them.



Handley/Schulzrinne/Schooler/Rosenberg                       [Page 61]

Internet Draft                    SIP                      July 13, 2000


   Even if the "display-name" is empty, the "name-addr" form MUST be
   used if the "addr-spec" contains a comma, question mark, or
   semicolon.  Syntax issues are discussed in Section 6.6.



        From            =  ( "From" | "f" ) ":" ( name-addr | addr-spec )
                           [ ";" tag-param ] *( ";" addr-extension )
        tag-param       =  "tag=" token
        addr-extension  =  generic-param


   Examples:


     From: "A. G. Bell"  ;tag=a48s
     From: sip:+12125551212@server.phone2net.com
     From: Anonymous 



   The "tag" MAY appear in the From field of a request. It MUST be
   present when it is possible that two instances of a user sharing a
   SIP address can make call invitations with the same Call-ID.

   The "tag" value MUST be globally unique and cryptographically random
   with at least 32 bits of randomness. A single user agent maintains
   the same tag at least within the same Call-ID, but it is RECOMMENDED
   to maintain the same tag across calls and instances of the UA
   application to allow restarting of a user agent.

   For the purpose of identifying call legs, two From or To header
   fields are equal if and only if:

        o The addr-spec component is equal, according to the rules in
          Section 2.1.

        o Any "tag" and addr-extension parameters are equal. Parameter
          names are compared as case-insensitive strings, while the
          parameter value is compared as a case-sensitive string. The
          tag comparison is only performed if both header fields have a
          tag value.


        Call-ID, To and From are needed to identify a call leg.
        The distinction between call and call leg matters in calls
        with multiple responses to a forked request. The format is
        similar to the equivalent RFC 822 [28] header, but with a



Handley/Schulzrinne/Schooler/Rosenberg                       [Page 62]

Internet Draft                    SIP                      July 13, 2000


        URI instead of just an email address.

6.26 Hide

   A client uses the Hide request header field to indicate that it wants
   the path comprised of the Via header fields (Section 6.47) to be
   hidden from subsequent proxies and user agents. It can take two
   forms: Hide: route and Hide:  hop. Hide header fields are typically
   added by the client user agent, but MAY be added by any proxy along
   the path.

   If a request contains the "Hide: route" header field, all following
   proxies SHOULD hide their previous hop. If a request contains the
   "Hide: hop" header field, only the next proxy SHOULD hide the
   previous hop and then remove the Hide option unless it also wants to
   remain anonymous.

   A server hides the previous hop by encrypting the "host" (in both
   sent-by and via-received), "port", "maddr" parts of the top-most Via
   header field with an algorithm of its choice. Servers SHOULD add
   additional "salt" to the "host" and "port" information prior to
   encryption to prevent malicious downstream proxies from guessing
   earlier parts of the path based on seeing identical encrypted Via
   headers.  Hidden Via fields are marked with the "hidden" Via option,
   as described in Section 6.47.

   A server that is capable of hiding Via headers MUST attempt to
   decrypt all Via headers marked as "hidden" to perform loop detection.
   Servers that are not capable of hiding can ignore hidden Via fields
   in their loop detection algorithm.


        If hidden headers were not marked, a proxy would have to
        decrypt all headers to detect loops, just in case one was
        encrypted, as the Hide: hop option may have been removed
        along the way.

   A client requesting "Hide: route" can only rely on keeping the
   request path private if it sends the request to a trusted proxy.
   Hiding the route of a SIP request is of limited value if the request
   results in data packets being exchanged directly between two user
   agents.

   The use of Hide header fields is discouraged unless path privacy is
   truly needed as Hide fields impose extra processing costs proxies.

   The encryption of Via header fields is described in more detail in
   Section 13.



Handley/Schulzrinne/Schooler/Rosenberg                       [Page 63]

Internet Draft                    SIP                      July 13, 2000


   The Hide header field has the following syntax:


        Hide        =  "Hide" ":" ( "route" | "hop" | other-hide )
        other-hide  =  token


6.27 In-Reply-To

   The In-Reply-To request header field enumerates the call-IDs that
   this call references or returns.


        This allows automatic call distribution systems to route
        return calls to the originator of the first call and allows
        callees to filter calls, so that only calls that return
        calls they have originated will be accepted. This field is
        not a substitute for request authentication.



        In-Reply-To  =  "In-Reply-To" ":" 1# callid


   Example:

   In-Reply-To: 70710@saturn.bell-tel.com, 17320@saturn.bell-tel.com



6.28 Max-Forwards

   The Max-Forwards request-header field may be used with any SIP method
   to limit the number of proxies or gateways that can forward the
   request to the next downstream server. This can also be useful when
   the client is attempting to trace a request chain which appears to be
   failing or looping in mid-chain.



        Max-Forwards  =  "Max-Forwards" ":" 1*DIGIT


   The Max-Forwards value is a decimal integer indicating the remaining
   number of times this request message is allowed to be forwarded.

   Each proxy or gateway recipient of a request containing a Max-
   Forwards header field MUST check and update its value prior to



Handley/Schulzrinne/Schooler/Rosenberg                       [Page 64]

Internet Draft                    SIP                      July 13, 2000


   forwarding the request. If the received value is zero (0), the
   recipient MUST NOT forward the request and returns 483 (Too many
   hops). Instead, a server MAY act as a final recipient for OPTIONS
   requests. It is RECOMMENDED that the server include Supported, Server
   and Allow header fields in the response.

   If the received Max-Forwards value is greater than zero, then the
   forwarded message MUST contain an updated Max-Forwards field with a
   value decremented by one (1).

   Example:

     Max-Forwards: 6



6.29 MIME-Version

   See [H19.4.1].

6.30 Organization

   The Organization general-header field conveys the name of the
   organization to which the entity issuing the request or response
   belongs. It MAY also be inserted by proxies at the boundary of an
   organization.


        The field MAY be used by client software to filter calls.



        Organization  =  "Organization" ":" TEXT-UTF8-TRIM


6.31 Priority

   The Priority request-header field indicates the urgency of the
   request as perceived by the client.



        Priority        =  "Priority" ":" priority-value
        priority-value  =  "emergency" | "urgent" | "normal"
                        |  "non-urgent" | other-priority
        other-priority     token





Handley/Schulzrinne/Schooler/Rosenberg                       [Page 65]

Internet Draft                    SIP                      July 13, 2000


   It is RECOMMENDED that the value of "emergency" only be used when
   life, limb or property are in imminent danger.

   Examples:


     Subject: A tornado is heading our way!
     Priority: emergency

     Subject: Weekend plans
     Priority: non-urgent




        These are the values of RFC 2076 [37], with the addition of
        "emergency".

6.32 Proxy-Authenticate

   The Proxy-Authenticate response-header field MUST be included as part
   of a 407 (Proxy Authentication Required) response.  It may also occur
   in a 401 (Unauthorized) response if the request was forked.  The
   field value consists of a challenge that indicates the authentication
   scheme and parameters applicable to the proxy for this Request-URI.

   Unlike its usage within HTTP, the Proxy-Authenticate header MUST be
   passed upstream in the response to the UAC. In SIP, only UAC's can
   authenticate themselves to proxies.

   The syntax for this header is defined in [H14.33]. See 14 for further
   details on its usage.

   A client SHOULD cache the credentials used for a particular proxy
   server and realm for the next request to that server. Credentials
   are, in general, valid for a specific value of the Request-URI at a
   particular proxy server. If a client contacts a proxy server that has
   required authentication in the past, but the client does not have
   credentials for the particular Request-URI, it MAY attempt to use the
   most-recently used credential. The server responds with 401
   (Unauthorized) if the client guessed wrong.


        This suggested caching behavior is motivated by proxies
        restricting phone calls to authenticated users. It seems
        likely that in most cases, all destinations require the
        same password. Note that end-to-end authentication is
        likely to be destination-specific.



Handley/Schulzrinne/Schooler/Rosenberg                       [Page 66]

Internet Draft                    SIP                      July 13, 2000


6.33 Proxy-Authorization

   The Proxy-Authorization request-header field allows the client to
   identify itself (or its user) to a proxy which requires
   authentication. The Proxy-Authorization field value consists of
   credentials containing the authentication information of the user
   agent for the proxy and/or realm of the resource being requested.

   Unlike Authorization, the Proxy-Authorization header field applies
   only to the next outbound proxy that demanded authentication using
   the Proxy- Authenticate field. When multiple proxies are used in a
   chain, the Proxy-Authorization header field is consumed by the first
   outbound proxy that was expecting to receive credentials. A proxy MAY
   relay the credentials from the client request to the next proxy if
   that is the mechanism by which the proxies cooperatively authenticate
   a given request.

   See [H14.34] for a definition of the syntax, and section 14 for a
   discussion of its usage.

6.34 Proxy-Require

   The Proxy-Require header field is used to indicate proxy-sensitive
   features that MUST be supported by the proxy. Any Proxy-Require
   header field features that are not supported by the proxy MUST be
   negatively acknowledged by the proxy to the client if not supported.
   Proxy servers treat this field identically to the Require field.

   See Section 6.36 for more details on the mechanics of this message
   and a usage example.

6.35 Record-Route

6.35.1 Operation

   The Record-Route request and response header field is added to a
   request by any proxy that insists on being in the path of subsequent
   requests for the same call leg.  A proxy SHOULD add it to any request
   for robustness, but a request route, once established, persists until
   the end of the call leg, regardless of whether the Record-Route
   header is present in subsequent requests.

   The Record-Route header field contains a globally reachable Request-
   URI that identifies the proxy server, including an address parameter
   that identifies its location. Each such proxy server adds the
   Request-URI of the incoming request to the beginning of the list.
   Requests between both user agents involved in the call leg, in either
   direction, traverse this route.



Handley/Schulzrinne/Schooler/Rosenberg                       [Page 67]

Internet Draft                    SIP                      July 13, 2000


        Some proxies, such as those controlling firewalls or in an
        automatic call distribution (ACD) system, need to maintain
        call state and thus need to receive any BYE, re-INVITE and
        ACK packets for the call. Note that proxy servers have to
        add Record-Route headers to each request as long as they
        want to be "visited" by the next request for the call leg.

   Proxies MUST include an maddr parameter in the URI in the Record-
   Route header, but MUST NOT include a transport parameter.

        If it is important that all requests go to the same host,
        server administrators are advised to be careful in
        selecting the appropriate name or address to ensure that
        name resolution does indeed resolve to the same host. For
        example, a domain name having an SRV record may resolve to
        a different network addresses on each attempt. Inclusion
        of, say, a TCP transport parameter may prevent a UA that
        supports only UDP but reached the proxy inserting the
        Record-Route via another proxy from reaching this proxy.

   The UAS copies the Record-Route request header field unchanged into
   the response. (Record-Route is only relevant for 2xx responses and
   responses where the server can expect the client to retry for the
   same Call-Id, as in 401 (Unauthorized) or 484 (Address Incomplete).)

6.35.2 Construction of Route Header

   Once a proxy P inserts a Record-Route header in a request from UA A
   to UA B, all subsequent requests from A to B and from B to A visit P.

   A UA builds the Route header field for subsequent requests from the
   Record-Route header fields received in either a response or a
   request.

   If a UAC finds a Record-Route header in a response, it copies it into
   Route header fields of all subsequent requests within the same call
   leg, reversing the order of fields, so that the first entry is the
   server closest to the UAC. If the response contained a Contact header
   field, the user agent adds its content as the last Route header.

   If a UA find a Record-Route header in a request, it copies the
   Record-Route maddr parameters only, maintaining their ordering, to
   the Route header field of future requests. Since the URIs contained
   in the Record-Route header fields are not useful for the reverse
   request path, the UA fills all other components of the Route name-
   addr value with the originating name-addr value.

   The originating name-addr is the name-addr value found in the Contact



Handley/Schulzrinne/Schooler/Rosenberg                       [Page 68]

Internet Draft                    SIP                      July 13, 2000


   header of the request or the From header field, if there is no
   Contact header field.

   If the request featured a Contact header field, the Contact header
   value is appended to the Route header list.

6.35.3 Request Destination

   Unless this would cause a loop, any client, including the UAC, SHOULD
   send the next request for this call leg to the first Request-URI in
   the Route request header field. A client MAY forward the request to a
   designated proxy instead, for example, if it lacks DNS resolution
   capability. If a client uses the first Route entry to route the
   request, it removes it.

6.35.4 Syntax

   The Record-Route header field has the following syntax:


        Record-Route  =  "Record-Route" ":" 1# name-addr [ rr-extension]
        rr-extension  =  token [ "=" ( token | quoted-string ) ]


   Proxy servers MUST include their address in a "maddr" URL parameter
   to ensure that subsequent requests are guaranteed to reach exactly
   the same server.

6.35.5 Example

   Example for a request where the proxy servers ieee.org and bell-
   telephone.com , in that order, insist on being part of subsequent
   request paths:

     Record-Route: ,
       



6.36 Require

   The Require general-header field is used by clients to tell user
   agent servers about options that the client expects the server to
   support in order to properly process the request. If a server does
   not understand the option, it MUST respond by returning status code
   420 (Bad Extension) and list those options it does not understand in
   the Unsupported header.




Handley/Schulzrinne/Schooler/Rosenberg                       [Page 69]

Internet Draft                    SIP                      July 13, 2000


        Require  =  "Require" ":" 1#option-tag


   Example:

   C->S:   INVITE sip:watson@bell-telephone.com SIP/2.0
           Require: com.example.billing
           Payment: sheep_skins, conch_shells

   S->C:   SIP/2.0 420 Bad Extension
           Unsupported: com.example.billing




        This is to make sure that the client-server interaction
        will proceed without delay when all options are understood
        by both sides, and only slow down if options are not
        understood (as in the example above).  For a well-matched
        client-server pair, the interaction proceeds quickly,
        saving a round-trip often required by negotiation
        mechanisms. In addition, it also removes ambiguity when the
        client requires features that the server does not
        understand. Some features, such as call handling fields,
        are only of interest to end systems.

   Proxy and redirect servers MUST ignore features that are not
   understood. If a particular extension requires that intermediate
   devices support it, the extension MUST be tagged in the Proxy-Require
   field as well (see Section 6.34).

6.37 Response-Key

   The Response-Key request-header field can be used by a client to
   request the key that the called user agent SHOULD use to encrypt the
   response with. The syntax is:



        Response-Key  =  "Response-Key" ":" key-scheme 1*SP #key-param
        key-scheme    =  token
        key-param     =  generic-param


   The "key-scheme" gives the type of encryption to be used for the
   response. Section 13 describes security schemes.

   If the client insists that the server return an encrypted response,



Handley/Schulzrinne/Schooler/Rosenberg                       [Page 70]

Internet Draft                    SIP                      July 13, 2000


   it includes a

                  Require: org.ietf.sip.encrypt-response

   header field in its request. If the server cannot encrypt for
   whatever reason, it MUST follow normal Require header field
   procedures and return a 420 (Bad Extension) response. If this Require
   header field is not present, a server SHOULD still encrypt if it can.

6.38 Retry-After

   The Retry-After response-header field can be used with a 503 (Service
   Unavailable) response to indicate how long the service is expected to
   be unavailable to the requesting client and with a 404 (Not Found),
   600 (Busy), or 603 (Decline) response to indicate when the called
   party anticipates being available again. The value of this field can
   be either an SIP-date or an integer number of seconds (in decimal)
   after the time of the response.

   A REGISTER request MAY include this header field when deleting
   registrations with "Contact: * ;expires: 0". The Retry-After value
   then indicates when the user might again be reachable. The registrar
   MAY then include this information in responses to future calls.

   An optional comment can be used to indicate additional information
   about the time of callback. An optional "duration" parameter
   indicates how long the called party will be reachable starting at the
   initial time of availability. If no duration parameter is given, the
   service is assumed to be available indefinitely.



        Retry-After  =  "Retry-After" ":" ( SIP-date | delta-seconds )
                        [ comment ] [ ";" "duration" "=" delta-seconds ]


   Examples of its use are

     Retry-After: Mon, 21 Jul 1997 18:48:34 GMT (I'm in a meeting)
     Retry-After: Mon, 01 Jan 9999 00:00:00 GMT
       (Dear John: Don't call me back, ever)
     Retry-After: Fri, 26 Sep 1997 21:00:00 GMT;duration=3600
     Retry-After: 120



   In the third example, the callee is reachable for one hour starting
   at 21:00 GMT. In the last example, the delay is 2 minutes.



Handley/Schulzrinne/Schooler/Rosenberg                       [Page 71]

Internet Draft                    SIP                      July 13, 2000


6.39 Route

   The Route request-header field determines the route taken by a
   request. Each host removes the first entry and then proxies the
   request to the host listed in that entry, also using it as the
   Request-URI. The operation is described in more detail in Section
   6.35.

   The Route header field has the following syntax:


        Route            =  "Route" ":" 1# name-addr [ route-extension ]
        route-extension  =  generic-param


6.40 Server

   The Server response-header field contains information about the
   software used by the user agent server to handle the request. The
   syntax for this field is defined in [H14.38].

6.41 Subject

   This header field provides a summary or indicates the nature of the
   call, allowing call filtering without having to parse the session
   description. (Note that the session description does not have to use
   the same subject indication as the invitation.)



        Subject  =  ( "Subject" | "s" ) ":" TEXT-UTF8-TRIM


   Example:


     Subject: Tune in - they are talking about your work!



6.42 Supported

   The Supported general-header field enumerates all the capabilities of
   the client or server. This header field SHOULD be included in all
   requests (except ACK) and in all responses.


        Including the header field in all responses greatly



Handley/Schulzrinne/Schooler/Rosenberg                       [Page 72]

Internet Draft                    SIP                      July 13, 2000


        simplifies the use of extensions for call control in
        subsequent transactions with the same server.

   Syntax:


        Supported  =  ( "Supported" | "k" ) ":" 1#option-tag


6.43 Timestamp

   The Timestamp general-header field describes when the client sent the
   request to the server. The value of the timestamp is of significance
   only to the client and it MAY use any timescale. The server MUST echo
   the exact same value and MAY, if it has accurate information about
   this, add a floating point number indicating the number of seconds
   that have elapsed since it has received the request.  The timestamp
   is used by the client to compute the round-trip time to the server so
   that it can adjust the timeout value for retransmissions.



        Timestamp  =  "Timestamp" ":" *(DIGIT) [ "." *(DIGIT) ] [ delay ]
        delay      =  *(DIGIT) [ "." *(DIGIT) ]


   Note that there MUST NOT be any LWS between a DIGIT and the decimal
   point.

6.44 To

   The To general-header field specifies recipient of the request.



        To  =  ( "To" | "t" ) ":" ( name-addr | addr-spec )
               [ ";" tag-param ] *( ";" addr-extension )


   Requests and responses MUST contain a To general-header field,
   indicating the desired recipient of the request. The optional
   "display-name" is meant to be rendered by a human-user interface. The
   UAS or redirect server copies the To header field into its response,
   and MUST add a "tag" parameter.


        If there was more than one Via header field, the request
        was handled by at least one proxy server. Since the



Handley/Schulzrinne/Schooler/Rosenberg                       [Page 73]

Internet Draft                    SIP                      July 13, 2000


        receiver cannot know whether any of the proxy servers
        forked the request, it is safest to assume that they might
        have.

   The SIP-URL MUST NOT contain the "transport-param", "maddr-param",
   "ttl-param", or "headers" elements. A server that receives a SIP-URL
   with these elements removes them before further processing.

   The "tag" parameter serves as a general mechanism to distinguish
   multiple instances of a user identified by a single SIP URL. As
   proxies can fork requests, the same request can reach multiple
   instances of a user (mobile and home phones, for example). As each
   can respond, there needs to be a means to distinguish the responses
   from each at the caller. The situation also arises with multicast
   requests. The tag in the To header field serves to distinguish
   responses at the UAC. It MUST be placed in the To field of the
   response by user agent, registrar and redirect servers, but MUST NOT
   be inserted into responses forwarded upstream by proxies. However,
   responses generated locally by a proxy, and then sent upstream, MUST
   contain a tag.

   A UAS or redirect server MUST add a "tag" parameter for all final
   responses for all transactions within a call leg. All such parameters
   have the same value within the same call leg. These servers MAY add
   the tag for informational responses during the initial INVITE
   transaction, but MUST add a tag to informational responses for all
   subsequent transactions.

   See Section 6.25 for details of the "tag" parameter. The "tag"
   parameter in To headers is ignored when matching responses to
   requests that did not contain a "tag" in their To header.

   Section 11 describes when the "tag" parameter MUST appear in
   subsequent requests. Note that if a request already contained a tag,
   this tag MUST be mirrored in the response; a new tag MUST NOT be
   inserted.

   Section 6.25 describes how To and From header fields are compared for
   the purpose of matching requests to call legs.

   UAS SHOULD accept requests even if they do not recognize the URI
   scheme (e.g., a tel: URI) or if the To header does not address the
   user. Only the Request-URI should be used to reject requests.

   Even if the "display-name" is empty, the "name-addr" form MUST be
   used if the "addr-spec" contains a comma, question mark, or
   semicolon.  Note that LWS is common, but not mandatory between the
   display-name and the "<".



Handley/Schulzrinne/Schooler/Rosenberg                       [Page 74]

Internet Draft                    SIP                      July 13, 2000


   The following are examples of valid To headers:

     To: The Operator ;tag=287447
     To: sip:+12125551212@server.phone2net.com




        Call-ID, To and From are needed to identify a call leg.
        The distinction between call and call leg matters in calls
        with multiple responses from a forked request. The "tag" is
        added to the To header field in the response to allow
        forking of future requests for the same call by proxies,
        while addressing only one of the possibly several
        responding user agent servers. It also allows several
        instances of the callee to send requests that can be
        distinguished.

6.45 Unsupported

   The Unsupported response-header field lists the features not
   supported by the server. See Section 6.36 for a usage example and
   motivation.

   Syntax:


        Unsupported  =  "Unsupported" ":" 1#option-tag


6.46 User-Agent

   The User-Agent general-header field contains information about the
   client user agent originating the request. The syntax and semantics
   are defined in [H14.42].

6.47 Via

   The Via field indicates the path taken by the request so far.  This
   prevents request looping and ensures replies take the same path as
   the requests, which assists in firewall traversal and other unusual
   routing situations.

6.47.1 Requests

   The client originating the request MUST insert into the request a Via
   field containing its host name or network address and, if not the
   default port number, the port number at which it wishes to receive



Handley/Schulzrinne/Schooler/Rosenberg                       [Page 75]

Internet Draft                    SIP                      July 13, 2000


   responses. (Note that this port number can differ from the UDP source
   port number of the request.) A fully-qualified domain name is
   RECOMMENDED. Each subsequent proxy server that sends the request
   onwards MUST add its own additional Via field before any existing Via
   fields. A proxy that receives a redirection (3xx) response and then
   searches recursively, MUST use the same Via headers as on the
   original proxied request.

   A proxy SHOULD check the top-most Via header field to ensure that it
   contains the sender's correct network address, as seen from that
   proxy. If the sender's address is incorrect, the proxy MUST add an
   additional "received" attribute, as described 6.47.2.


        A host behind a network address translator (NAT) or
        firewall may not be able to insert a network address into
        the Via header that can be reached by the next hop beyond
        the NAT. Use of the received attribute allows SIP requests
        to traverse NAT's which only modify the source IP address.
        NAT's which modify port numbers, called Network Address
        Port Translators (NAPTs) will not properly pass SIP when
        transported on UDP, so that an application-layer gateway is
        required. When run over TCP, SIP stands a better chance of
        traversing NAPTs, since its behavior is similar to HTTP in
        this case, albeit using different ports.

   A client that sends a request to a multicast address MUST add the
   "maddr" parameter to its Via header field, and SHOULD add the "ttl"
   parameter. (In that case, the maddr parameter SHOULD contain the
   destination multicast address, although under exceptional
   circumstances it MAY contain a unicast address.) If a server receives
   a request which contained an "maddr" parameter in the topmost Via
   field, it SHOULD send the response to the address listed in the
   "maddr" parameter.

   Loop detection is described in Section 12.3.1.

6.47.2 Receiver-tagged Via Header Fields

   Every host that sends or forwards a SIP request adds a Via field
   indicating the host's address. However, it is possible that Network
   Address Translators (NATs) change the source address and port of the
   request (e.g., from a net-10 to a globally routable address), in
   which case the Via header field cannot be relied on to route replies.
   To prevent this, a proxy SHOULD check the top-most Via header field
   to ensure that it contains the sender's correct network address, as
   seen from that proxy. If the sender's address is incorrect, the proxy
   MUST add a "received" parameter to the Via header field inserted by



Handley/Schulzrinne/Schooler/Rosenberg                       [Page 76]

Internet Draft                    SIP                      July 13, 2000


   the previous hop. Such a modified Via header field is known as a
   receiver-tagged Via header field.  If and only if the source address
   and port differ from the sent-by address, the proxy also includes the
   source port in the received parameter.

   An example is:


     Via: SIP/2.0/UDP erlang.bell-telephone.com:5060
     Via: SIP/2.0/UDP 10.0.0.1:5060 ;received=199.172.136.3



   In this example, the message originated from 10.0.0.1 and traversed a
   NAT with the external address border.ieee.org (199.172.136.3) to
   reach erlang.bell-telephone.com.  The latter noticed the mismatch,
   and added a parameter to the previous hop's Via header field,
   containing the address that the packet actually came from. (Note that
   the NAT border.ieee.org is not a SIP server.)

6.47.3 Responses

   Via header fields in responses are processed by a proxy or UAC
   according to the following rules:

        1.   The first Via header field should indicate the proxy or
             client processing this response. If it does not, discard
             the message.  Otherwise, remove this Via field.

        2.   If there is no second Via header field, this response is
             destined for this client. Otherwise, the processing depends
             on whether the Via field contains a "maddr" parameter or is
             a receiver-tagged field:

             - If the second Via header field contains a "maddr"
               parameter, forward the response to the address listed
               there, using the port indicated in "sent-by", or port
               5060 if none is present. If the address is a multicast
               address, the response SHOULD be sent using the TTL
               indicated in the "ttl" parameter, or with a TTL of 1 if
               that parameter is not present.

             - If the second Via header field does not contain a "maddr"
               parameter and is a receiver-tagged field (Section
               6.47.2), send the message to the address in the
               "received" parameter, using the port indicated in the
               "sent-by" value, or using port 5060 if none is present.




Handley/Schulzrinne/Schooler/Rosenberg                       [Page 77]

Internet Draft                    SIP                      July 13, 2000


             - If neither of the previous cases apply, send the message
               to the address indicated by the "sent-by" value in the
               second Via header field.

6.47.4 User Agent and Redirect Servers

   A UAS or redirect server copies the Via header fields into the
   response, without changing their order, and then sends a response
   based on one of the following rules:

        o If the first Via header field in the request contains a
          "maddr" parameter, send the response to the address listed
          there, using the port indicated in "sent-by", or port 5060 if
          none is present. If the address is a multicast address, the
          response SHOULD be sent using the TTL indicated in the "ttl"
          parameter, or with a TTL of 1 if that parameter is not
          present.

        o If the address in the "sent-by" value of the first Via field
          differs from the source address of the packet, send the
          response to the actual packet source address, similar to the
          treatment for receiver-tagged Via header fields (Section
          6.47.2). However, the port number is taken from the sent-by
          part of the Via header.


             This mode of operation supports unaided traversal of
             SIP responses through NATs, but does not work through
             NAPTs.

        o If neither of these conditions is true, send the response to
          the address contained in the "sent-by" value. If the request
          was sent using TCP, use the existing TCP connection if
          available.

6.47.5 Syntax

   The format for a Via header field is shown in Fig. 10. The "maddr"
   parameter, designating the multicast address, and the "ttl"
   parameter, designating the time-to-live (TTL) value, are included
   only if the request was sent via multicast. The "received" parameter
   is added only for receiver-added Via fields (Section 6.47.2).  For
   reasons of privacy, a client or proxy may wish to hide its Via
   information by encrypting it (see Section 6.26).  The "hidden"
   parameter is included if this header field was hidden by the upstream
   proxy (see 6.26). Note that privacy of the proxy relies on the
   cooperation of the next hop, as the next-hop proxy will, by
   necessity, know the IP address and port number of the source host.



Handley/Schulzrinne/Schooler/Rosenberg                       [Page 78]

Internet Draft                    SIP                      July 13, 2000




  Via              = ( "Via" | "v") ":" 1#( sent-protocol sent-by
                     *( ";" via-params ) [ comment ] )
  via-params       = via-hidden | via-ttl | via-maddr 
                   | via-received | via-branch | via-extension
  via-hidden       = "hidden"
  via-ttl          = "ttl" "=" ttl
  via-maddr        = "maddr" "=" maddr
  via-received	   = "received" "=" host [ ":" port ]
  via-branch       = "branch" "=" token
  via-extension    = token [ "=" ( token | quoted-string ) ]
  sent-protocol    = protocol-name "/" protocol-version "/" transport
  protocol-name    = "SIP" | token
  protocol-version = token
  transport        = "UDP" | "TCP" | token
  sent-by          = ( host [ ":" port ] ) | ( concealed-host )
  concealed-host   = token
  ttl              = 1*3DIGIT     ; 0 to 255


   Figure 10: Syntax of Via header field



   The "branch" parameter is included by every proxy. The token MUST be
   unique for each distinct request. The precise format of the token is
   implementation-defined. In order to be able to both detect loops and
   associate responses with the corresponding request, the parameter
   SHOULD consist of two parts separable by the implementation. One
   part, used for loop detection (Section 12.3.1), MAY be computed as a
   cryptographic hash of the To, From, Call-ID header fields, the
   Request-URI of the request received (before translation) and the
   sequence number from the CSeq header field. The algorithm used to
   compute the hash is implementation-dependent, but MD5 [38], expressed
   in hexadecimal, is a reasonable choice. (Note that base64 is not
   permissible for a token.) The other part, used for matching responses
   to requests, is a function of the branch taken, for example, a
   sequence number or the MD5 hash of the request-URI of the request
   sent on the branch.

   For example: 7a83e5750418bce23d5106b4c06cc632.1


        The "branch" parameter MUST depend on the incoming
        request-URI to distinguish looped requests from requests
        whose request-URI is changed and which then reach a server
        visited earlier.



Handley/Schulzrinne/Schooler/Rosenberg                       [Page 79]

Internet Draft                    SIP                      July 13, 2000


   CANCEL requests MUST have the same branch value as the corresponding
   forked request. When a response arrives at the proxy it can use the
   branch value to figure out which branch the response corresponds to.


     Via: SIP/2.0/UDP first.example.com:4000;ttl=16
       ;maddr=224.2.0.1 ;branch=a7c6a8dlze.1 (Acme server)
     Via: SIP/2.0/UDP adk8



6.48 Warning

   The Warning response-header field is used to carry additional
   information about the status of a response. Warning headers are sent
   with responses and have the following format:



        Warning        =  "Warning" ":" 1#warning-value
        warning-value  =  warn-code SP warn-agent SP warn-text
        warn-code      =  3DIGIT
        warn-agent     =  ( host [ ":" port ] ) | pseudonym
                          ;  the name or pseudonym of the server adding
                          ;  the Warning header, for use in debugging
        warn-text      =  quoted-string


   A response MAY carry more than one Warning header.

   The "warn-text" should be in a natural language that is most likely
   to be intelligible to the human user receiving the response.  This
   decision can be based on any available knowledge, such as the
   location of the cache or user, the Accept-Language field in a
   request, or the Content-Language field in a response. The default
   language is i-default [39].

   Any server MAY add Warning headers to a response. Proxy servers MUST
   place additional Warning headers before any Authorization headers.
   Within that constraint, Warning headers MUST be added after any
   existing Warning headers not covered by a signature. A proxy server
   MUST NOT delete any Warning header field that it received with a
   response.

   When multiple Warning headers are attached to a response, the user
   agent SHOULD display as many of them as possible, in the order that
   they appear in the response. If it is not possible to display all of
   the warnings, the user agent first displays warnings that appear



Handley/Schulzrinne/Schooler/Rosenberg                       [Page 80]

Internet Draft                    SIP                      July 13, 2000


   early in the response.

   The warn-code consists of three digits. A first digit of "3"
   indicates warnings specific to SIP.

   This is a list of the currently-defined "warn-code"s, each with a
   recommended warn-text in English, and a description of its meaning.
   Note that these warnings describe failures induced by the session
   description.

   Warnings 300 through 329 are reserved for indicating problems with
   keywords in the session description, 330 through 339 are warnings
   related to basic network services requested in the session
   description, 370 through 379 are warnings related to quantitative QoS
   parameters requested in the session description, and 390 through 399
   are miscellaneous warnings that do not fall into one of the above
   categories.

        300 Incompatible network protocol: One or more network protocols
             contained in the session description are not available.

        301 Incompatible network address formats: One or more network
             address formats contained in the session description are
             not available.

        302 Incompatible transport protocol: One or more transport
             protocols described in the session description are not
             available.

        303 Incompatible bandwidth units: One or more bandwidth
             measurement units contained in the session description were
             not understood.

        304 Media type not available: One or more media types contained
             in the session description are not available.

        305 Incompatible media format: One or more media formats
             contained in the session description are not available.

        306 Attribute not understood: One or more of the media
             attributes in the session description are not supported.

        307 Session description parameter not understood: A parameter
             other than those listed above was not understood.

        330 Multicast not available: The site where the user is located
             does not support multicast.




Handley/Schulzrinne/Schooler/Rosenberg                       [Page 81]

Internet Draft                    SIP                      July 13, 2000


        331 Unicast not available: The site where the user is located
             does not support unicast communication (usually due to the
             presence of a firewall).

        370 Insufficient bandwidth: The bandwidth specified in the
             session description or defined by the media exceeds that
             known to be available.

        399 Miscellaneous warning: The warning text can include
             arbitrary information to be presented to a human user, or
             logged. A system receiving this warning MUST NOT take any
             automated action.


        1xx and 2xx have been taken by HTTP/1.1.

   Additional "warn-code"s, as in the example below, can be defined
   through IANA.

   Examples:


     Warning: 307 isi.edu "Session parameter 'foo' not understood"
     Warning: 301 isi.edu "Incompatible network address type 'E.164'"



6.49 WWW-Authenticate

   The WWW-Authenticate response-header field MUST be included in 401
   (Unauthorized) response messages. The field value consists of at
   least one challenge that indicates the authentication scheme(s) and
   parameters applicable to the Request-URI. See [H14.46] for a
   definition of the syntax, and section 14 for an overview of usage.

   The content of the "realm" parameter SHOULD be displayed to the user.
   A user agent SHOULD cache the authorization credentials for a given
   value of the destination (To header) and "realm" and attempt to re-
   use these values on the next request for that destination.

   In addition to the "basic" and "digest" authentication schemes
   defined in the specifications cited above, SIP defines a new scheme,
   PGP (RFC 2015, [40]), Section 15. Other schemes, such as S/MIME, are
   for further study.

7 Status Code Definitions

   The response codes are consistent with, and extend, HTTP/1.1 response



Handley/Schulzrinne/Schooler/Rosenberg                       [Page 82]

Internet Draft                    SIP                      July 13, 2000


   codes. Not all HTTP/1.1 response codes are appropriate, and only
   those that are appropriate are given here. Other HTTP/1.1 response
   codes SHOULD NOT be used. Response codes not defined by HTTP/1.1 have
   codes x80 upwards to avoid clashes with future HTTP response codes.
   Also, SIP defines a new class, 6xx. The default behavior for unknown
   response codes is given for each category of codes.

7.1 Informational 1xx

   Informational responses indicate that the server or proxy contacted
   is performing some further action and does not yet have a definitive
   response. The client SHOULD wait for a further response from the
   server, and the server SHOULD send such a response without further
   prompting. A server SHOULD send a 1xx response if it expects to take
   more than 200 ms to obtain a final response. A server MAY issue zero
   or more 1xx responses, with no restriction on their ordering or
   uniqueness. Note that 1xx responses are not transmitted reliably,
   that is, they do not cause the client to send an ACK. Servers are
   free to retransmit informational responses and clients can inquire
   about the current state of call processing by re-sending the request.

7.1.1 100 Trying

   Some unspecified action is being taken on behalf of this call (e.g.,
   a database is being consulted), but the user has not yet been
   located.

7.1.2 180 Ringing

   The called user agent has located a possible location where the user
   has registered recently and is trying to alert the user.

7.1.3 181 Call Is Being Forwarded

   A proxy server MAY use this status code to indicate that the call is
   being forwarded to a different set of destinations.

7.1.4 182 Queued

   The called party is temporarily unavailable, but the callee has
   decided to queue the call rather than reject it. When the callee
   becomes available, it will return the appropriate final status
   response. The reason phrase MAY give further details about the status
   of the call, e.g., "5 calls queued; expected waiting time is 15
   minutes". The server MAY issue several 182 responses to update the
   caller about the status of the queued call.

7.1.5 183 Session Progress



Handley/Schulzrinne/Schooler/Rosenberg                       [Page 83]

Internet Draft                    SIP                      July 13, 2000


   The 183 (Session Progress) response is used to convey information
   about the progress of the call which is not otherwise classified. The
   Reason-Phrase MAY be used to convey more details about the call
   progress. The status response MAY contain a session description that
   allows the UAC to receive special announcements or progress tones.

7.2 Successful 2xx

   The request was successful and MUST terminate a search.

7.2.1 200 OK

   The request has succeeded. The information returned with the response
   depends on the method used in the request, for example:

        BYE: The call has been terminated. The message body is empty.

        CANCEL: The search has been cancelled. The message body is
             empty.

        INVITE: The callee has agreed to participate; the message body
             indicates the callee's capabilities.

        OPTIONS: The callee has agreed to share its capabilities,
             included in the message body.

        REGISTER: The registration has succeeded. The client treats the
             message body according to its Content-Type.

7.3 Redirection 3xx

   3xx responses give information about the user's new location, or
   about alternative services that might be able to satisfy the call.
   They SHOULD terminate an existing search, and MAY cause the initiator
   to begin a new search if appropriate.

   To avoid forwarding loops, a user agent client or proxy MUST check
   whether the address returned by a redirect server equals an address
   tried earlier.

7.3.1 300 Multiple Choices

   The address in the request resolved to several choices, each with its
   own specific location, and the user (or user agent) can select a
   preferred communication end point and redirect its request to that
   location.

   The response SHOULD include an entity containing a list of resource



Handley/Schulzrinne/Schooler/Rosenberg                       [Page 84]

Internet Draft                    SIP                      July 13, 2000


   characteristics and location(s) from which the user or user agent can
   choose the one most appropriate, if allowed by the Accept request
   header. The entity format is specified by the media type given in the
   Content-Type header field. The choices SHOULD also be listed as
   Contact fields (Section 6.15).  Unlike HTTP, the SIP response MAY
   contain several Contact fields or a list of addresses in a Contact
   field. User agents MAY use the Contact header field value for
   automatic redirection or MAY ask the user to confirm a choice.
   However, this specification does not define any standard for such
   automatic selection.


        This status response is appropriate if the callee can be
        reached at several different locations and the server
        cannot or prefers not to proxy the request.

7.3.2 301 Moved Permanently

   The user can no longer be found at the address in the Request-URI and
   the requesting client SHOULD retry at the new address given by the
   Contact header field (Section 6.15). The caller SHOULD update any
   local directories, address books and user location caches with this
   new value and redirect future requests to the address(es) listed.

7.3.3 302 Moved Temporarily

   The requesting client SHOULD retry the request at the new address(es)
   given by the Contact header field (Section 6.15).  The Request-URI of
   the new request uses the value of the Contact header in the response.
   The new request can take two different forms. In the first approach,
   the To, From, Call-ID, and CSeq header fields in the new request are
   the same as in the original request, with a new branch identifier in
   the Via header field. Proxies MUST follow this behavior and UACs MAY.
   UAs MAY also use the Contact information for the To header field, as
   well as a new Call-ID value.


        Reusing the CSeq value allows proxies to avoid forwarding
        the request to the same destination twice, as a proxy will
        consider it a retransmission.

   The duration of the redirection can be indicated through an Expires
   (Section 6.24) header. If there is no explicit expiration time, the
   address is only valid for this call and MUST NOT be cached for future
   calls.

7.3.4 305 Use Proxy




Handley/Schulzrinne/Schooler/Rosenberg                       [Page 85]

Internet Draft                    SIP                      July 13, 2000


   The requested resource MUST be accessed through the proxy given by
   the Contact field. The Contact field gives the URI of the proxy. The
   recipient is expected to repeat this single request via the proxy.
   305 responses MUST only be generated by user agent servers.

7.3.5 380 Alternative Service

   The call was not successful, but alternative services are possible.
   The alternative services are described in the message body of the
   response.  Formats for such bodies are not defined here, and may be
   the subject of future standardization.

7.4 Request Failure 4xx

   4xx responses are definite failure responses from a particular
   server.  The client SHOULD NOT retry the same request without
   modification (e.g., adding appropriate authorization). However, the
   same request to a different server might be successful.

7.4.1 400 Bad Request

   The request could not be understood due to malformed syntax.  The
   Reason-Phrase SHOULD identify the syntax problem in more detail,
   e.g., "Missing Content-Length header".

7.4.2 401 Unauthorized

   The request requires user authentication.  This response is issued by
   user agent servers and registrars, while 407 (Proxy Authentication
   Required) is used by proxy servers.

7.4.3 402 Payment Required

   Reserved for future use.

7.4.4 403 Forbidden

   The server understood the request, but is refusing to fulfill it.
   Authorization will not help, and the request SHOULD NOT be repeated.

7.4.5 404 Not Found

   The server has definitive information that the user does not exist at
   the domain specified in the Request-URI. This status is also returned
   if the domain in the Request-URI does not match any of the domains
   handled by the recipient of the request.

7.4.6 405 Method Not Allowed



Handley/Schulzrinne/Schooler/Rosenberg                       [Page 86]

Internet Draft                    SIP                      July 13, 2000


   The method specified in the Request-Line is not allowed for the
   address identified by the Request-URI. The response MUST include an
   Allow header field containing a list of valid methods for the
   indicated address.

7.4.7 406 Not Acceptable

   The resource identified by the request is only capable of generating
   response entities which have content characteristics not acceptable
   according to the accept headers sent in the request.

7.4.8 407 Proxy Authentication Required

   This code is similar to 401 (Unauthorized), but indicates that the
   client MUST first authenticate itself with the proxy. The proxy MUST
   return a Proxy-Authenticate header field (section 6.32) containing a
   challenge applicable to the proxy for the requested resource. The
   client MAY repeat the request with a suitable Proxy-Authorization
   header field (section 6.33). SIP access authentication is explained
   in section 13.2 and 14.

   This status code is used for applications where access to the
   communication channel (e.g., a telephony gateway) rather than the
   callee requires authentication.

7.4.9 408 Request Timeout

   The server could not produce a response within a suitable amount of
   time, for example, since it could not determine the location of the
   user in time. The amount of time may have been indicated in the
   Expires request-header field or may be set by the server. The client
   MAY repeat the request without modifications at any later time.

7.4.10 409 Conflict

   The request could not be completed due to a conflict with the current
   state of the resource. This response is returned if the action
   parameter in a REGISTER request conflicts with existing
   registrations.

7.4.11 410 Gone

   The requested resource is no longer available at the server and no
   forwarding address is known. This condition is expected to be
   considered permanent. If the server does not know, or has no facility
   to determine, whether or not the condition is permanent, the status
   code 404 (Not Found) SHOULD be used instead.




Handley/Schulzrinne/Schooler/Rosenberg                       [Page 87]

Internet Draft                    SIP                      July 13, 2000


7.4.12 411 Length Required

   The server refuses to accept the request without a defined Content-
   Length. The client MAY repeat the request if it adds a valid
   Content-Length header field containing the length of the message-body
   in the request message.

7.4.13 413 Request Entity Too Large

   The server is refusing to process a request because the request
   entity is larger than the server is willing or able to process. The
   server MAY close the connection to prevent the client from continuing
   the request.

   If the condition is temporary, the server SHOULD include a Retry-
   After header field to indicate that it is temporary and after what
   time the client MAY try again.

7.4.14 414 Request-URI Too Long

   The server is refusing to service the request because the Request-URI
   is longer than the server is willing to interpret.

7.4.15 415 Unsupported Media Type

   The server is refusing to service the request because the message
   body of the request is in a format not supported by the server for
   the requested method. The server SHOULD return a list of acceptable
   formats using the Accept, Accept-Encoding and Accept-Language header
   fields.

7.4.16 420 Bad Extension

   The server did not understand the protocol extension specified in a
   Proxy-Require (Section 6.34) or Require (Section 6.36) header field.

7.4.17 480 Temporarily Unavailable

   The callee's end system was contacted successfully but the callee is
   currently unavailable (e.g., not logged in, logged in in such a
   manner as to preclude communication with the callee or activated the
   "do not disturb" feature). The response MAY indicate a better time to
   call in the Retry-After header. The user could also be available
   elsewhere (unbeknownst to this host), thus, this response does not
   terminate any searches. The reason phrase SHOULD indicate a more
   precise cause as to why the callee is unavailable. This value SHOULD
   be setable by the user agent. Status 486 (Busy Here) MAY be used to
   more precisely indicate a particular reason for the call failure.



Handley/Schulzrinne/Schooler/Rosenberg                       [Page 88]

Internet Draft                    SIP                      July 13, 2000


   This status is also returned by a redirect server that recognizes the
   user identified by the Request-URI, but does not currently have a
   valid forwarding location for that user.

7.4.18 481 Call Leg/Transaction Does Not Exist

   This status is returned under three conditions: The server received a
   BYE request that does not match any existing call leg, the server
   received a CANCEL request that does not match any existing
   transaction or the server received an INVITE with a To tag that does
   not match the local tag value. (A server simply discards an ACK
   referring to an unknown transaction.)

7.4.19 482 Loop Detected

   The server received a request with a Via (Section 6.47) path
   containing itself.

7.4.20 483 Too Many Hops

   The server received a request that contains a Max-Forwards (Section
   6.28) header with the value zero.

7.4.21 484 Address Incomplete

   The server received a request with a To (Section 6.44) address or
   Request-URI that was incomplete. Additional information SHOULD be
   provided.


        This status code allows overlapped dialing. With overlapped
        dialing, the client does not know the length of the dialing
        string. It sends strings of increasing lengths, prompting
        the user for more input, until it no longer receives a 484
        status response.

7.4.22 485 Ambiguous

   The callee address provided in the request was ambiguous. The
   response MAY contain a listing of possible unambiguous addresses in
   Contact headers.

   Revealing alternatives can infringe on privacy concerns of the user
   or the organization. It MUST be possible to configure a server to
   respond with status 404 (Not Found) or to suppress the listing of
   possible choices if the request address was ambiguous.

   Example response to a request with the URL lee@example.com :



Handley/Schulzrinne/Schooler/Rosenberg                       [Page 89]

Internet Draft                    SIP                      July 13, 2000


   485 Ambiguous SIP/2.0
   Contact: Carol Lee 
   Contact: Ping Lee 
   Contact: Lee M. Foote 




        Some email and voice mail systems provide this
        functionality. A status code separate from 3xx is used
        since the semantics are different: for 300, it is assumed
        that the same person or service will be reached by the
        choices provided. While an automated choice or sequential
        search makes sense for a 3xx response, user intervention is
        required for a 485 response.

7.4.23 486 Busy Here

   The callee's end system was contacted successfully but the callee is
   currently not willing or able to take additional calls at this end
   system. The response MAY indicate a better time to call in the
   Retry-After header. The user could also be available elsewhere, such
   as through a voice mail service, thus, this response does not
   terminate any searches. Status 600 (Busy Everywhere) SHOULD be used
   if the client knows that no other end system will be able to accept
   this call.

7.4.24 487 Request Terminated

   The original request was terminated by a BYE or CANCEL request.

7.4.25 488 Not Acceptable Here

   The response has the same meaning as 606 (Not Acceptable), but only
   applies to the specific entity addressed by the Request-URI and the
   request may succeed elsewhere.

7.5 Server Failure 5xx

   5xx responses are failure responses given when a server itself has
   erred. They are not definitive failures, and MUST NOT terminate a
   search if other possible locations remain untried.

7.5.1 500 Server Internal Error

   The server encountered an unexpected condition that prevented it from
   fulfilling the request. The client MAY display the specific error
   condition, and MAY retry the request after several seconds.



Handley/Schulzrinne/Schooler/Rosenberg                       [Page 90]

Internet Draft                    SIP                      July 13, 2000


   If the condition is temporary, the server MAY indicate when the
   client may retry the request using the Retry-After header.

7.5.2 501 Not Implemented

   The server does not support the functionality required to fulfill the
   request. This is the appropriate response when the server does not
   recognize the request method and is not capable of supporting it for
   any user.

7.5.3 502 Bad Gateway

   The server, while acting as a gateway or proxy, received an invalid
   response from the downstream server it accessed in attempting to
   fulfill the request.

7.5.4 503 Service Unavailable

   The server is currently unable to handle the request due to a
   temporary overloading or maintenance of the server. The implication
   is that this is a temporary condition which will be alleviated after
   some delay. If known, the length of the delay MAY be indicated in a
   Retry-After header. If no Retry-After is given, the client MUST
   handle the response as it would for a 500 response.

   Note: The existence of the 503 status code does not imply that a
   server has to use it when becoming overloaded. Some servers MAY wish
   to simply refuse the connection.

7.5.5 504 Server Time-out

   The server did not receive a timely response from the server (e.g., a
   location server) it accessed in attempting to process the request.
   Note that 408 (Request Timeout) should be used if there was no
   response within the period specified in the Expires header field from
   the upstream server.

7.5.6 505 Version Not Supported

   The server does not support, or refuses to support, the SIP protocol
   version that was used in the request message. The server is
   indicating that it is unable or unwilling to complete the request
   using the same major version as the client, other than with this
   error message. The response MAY contain an entity describing why that
   version is not supported and what other protocols are supported by
   that server. The format for such an entity is not defined here and
   may be the subject of future standardization.




Handley/Schulzrinne/Schooler/Rosenberg                       [Page 91]

Internet Draft                    SIP                      July 13, 2000


7.6 Global Failures 6xx

   6xx responses indicate that a server has definitive information about
   a particular user, not just the particular instance indicated in the
   Request-URI. All further searches for this user are doomed to failure
   and pending searches SHOULD be terminated.

7.6.1 600 Busy Everywhere

   The callee's end system was contacted successfully but the callee is
   busy and does not wish to take the call at this time. The response
   MAY indicate a better time to call in the Retry-After header. If the
   callee does not wish to reveal the reason for declining the call, the
   callee uses status code 603 (Decline) instead. This status response
   is returned only if the client knows that no other end point (such as
   a voice mail system) will answer the request. Otherwise, 486 (Busy
   Here) should be returned.

7.6.2 603 Decline

   The callee's machine was successfully contacted but the user
   explicitly does not wish to or cannot participate. The response MAY
   indicate a better time to call in the Retry-After header.

7.6.3 604 Does Not Exist Anywhere

   The server has authoritative information that the user indicated in
   the To request field does not exist anywhere. Searching for the user
   elsewhere will not yield any results.

7.6.4 606 Not Acceptable

   The user's agent was contacted successfully but some aspects of the
   session description such as the requested media, bandwidth, or
   addressing style were not acceptable.

   A 606 (Not Acceptable) response means that the user wishes to
   communicate, but cannot adequately support the session described. The
   606 (Not Acceptable) response MAY contain a list of reasons in a
   Warning header field describing why the session described cannot be
   supported. Reasons are listed in Section 6.48.  It is hoped that
   negotiation will not frequently be needed, and when a new user is
   being invited to join an already existing conference, negotiation may
   not be possible. It is up to the invitation initiator to decide
   whether or not to act on a 606 (Not Acceptable) response.

8 SIP Message Body




Handley/Schulzrinne/Schooler/Rosenberg                       [Page 92]

Internet Draft                    SIP                      July 13, 2000


8.1 Body Inclusion

   Requests MAY contain message bodies unless otherwise noted. In this
   specification, the CANCEL request MUST NOT contain a message body.

   The use of message bodies for REGISTER requests is for further study.

   For response messages, the request method and the response status
   code determine the type and interpretation of any message body. All
   responses MAY include a body. Message bodies for 1xx responses
   contain advisory information about the progress of the request.  1xx
   responses to INVITE requests MAY contain session descriptions.  Their
   interpretation depends on the response status code, but generally
   informs the caller what kind of session the callee is likely to
   establish, subject to later modification in the 2xx response.
   Request methods not defined in this specification MAY also contain
   session descriptions.  2xx responses to INVITE requests contain
   session descriptions. In 3xx responses, the message body MAY contain
   the description of alternative destinations or services, as described
   in Section 7.3. For responses with status 400 or greater, the message
   body MAY contain additional, human-readable information about the
   reasons for failure. It is RECOMMENDED that information in 1xx and
   300 and greater responses be of type text/plain or text/html

8.2 Message Body Type

   The Internet media type of the message body MUST be given by the
   Content-Type header field. If the body has undergone any encoding
   (such as compression) then this MUST be indicated by the Content-
   Encoding header field, otherwise Content-Encoding MUST be omitted. If
   applicable, the character set of the message body is indicated as
   part of the Content-Type header-field value.

   The "multipart" MIME type [41] MAY be used within the body of the
   message. Clients that send requests containing multipart message
   bodies MUST be able to send a session description as a non-multipart
   message body if the server requests this through an Accept header
   field.

8.3 Message Body Length

   The body length in bytes SHOULD be given by the Content-Length header
   field. Section 6.19 describes the behavior in detail.

   The "chunked" transfer encoding of HTTP/1.1 MUST NOT be used for SIP.
   (Note: The chunked encoding modifies the body of a message in order
   to transfer it as a series of chunks, each with its own size
   indicator.)



Handley/Schulzrinne/Schooler/Rosenberg                       [Page 93]

Internet Draft                    SIP                      July 13, 2000


9 Compact Form

   When SIP is carried over UDP with authentication and a complex
   session description, it may be possible that the size of a request or
   response is larger than the MTU. To address this problem, a more
   compact form of SIP is also defined by using abbreviations for the
   common header fields listed below:


   short field name  long field name   note
   c                 Content-Type
   e                 Content-Encoding
   f                 From
   i                 Call-ID
   k                 Supported         from "know"
   m                 Contact           from "moved"
   l                 Content-Length
   s                 Subject
   t                 To
   v                 Via


   Thus, the message in section 16.2 could also be written:


     INVITE sip:bob@example.com SIP/2.0
     v:SIP/2.0/UDP 131.215.131.131;maddr=239.128.16.254;ttl=16
     v:SIP/2.0/UDP 216.112.6.38
     f:sip:alice@wonderland.com
     t:sip:bob@example.com
     m:sip:alice@mouse.wonderland.com
     i:62729-27@216.112.6.38
     c:application/sdp
     CSeq: 4711 INVITE
     l:187

     v=0
     o=user1 53655765 2353687637 IN IP4 128.3.4.5
     s=Mbone Audio
     i=Discussion of Mbone Engineering Issues
     e=mbone@somewhere.com
     c=IN IP4 224.2.0.1/127
     t=0 0
     m=audio 3456 RTP/AVP 0
     a=rtpmap:0 PCMU/8000






Handley/Schulzrinne/Schooler/Rosenberg                       [Page 94]

Internet Draft                    SIP                      July 13, 2000


   Clients MAY mix short field names and long field names within the
   same request. Servers MUST accept both short and long field names for
   requests. Proxies MAY change header fields between their long and
   short forms, but this MUST NOT be done to fields following an
   Authorization header.

10 Behavior of SIP Clients and Servers

10.1 General Remarks

   SIP is defined so it can use either UDP (unicast or multicast) or TCP
   as a transport protocol; it provides its own reliability mechanism.

10.1.1 Requests

   Servers discard isomorphic requests, but first retransmit the
   appropriate response. (SIP requests are said to be idempotent , i.e.,
   receiving more than one copy of a request does not change the server
   state.)

   An incoming request is accepted unless the Call-ID value matches an
   existing call and the request has a tag value of To header that does
   not match the user agent server's tag value. In that case, the
   request is refused with a 481 (Call Leg/Transaction Does Not Exist)
   response.

   If the request is accepted and matches an existing call leg, the
   server compares the CSeq header field value. If less than or equal to
   the current sequence number, the request is a retransmission.
   Otherwise, it is a new request. If the request does not match an
   existing call leg, a new call leg is created.

   If the request did not contain a To tag value, the server returns a
   response containing the same To header field value as in the request
   and adds a unique tag.

   After receiving a CANCEL request from an upstream client, a stateful
   proxy server MAY send a CANCEL on all branches where it has not yet
   received a final response.

10.1.2 Responses

   A server MAY issue one or more provisional responses at any time
   before sending a final response. If a stateful proxy, user agent
   server, redirect server or registrar cannot respond to a request with
   a final response within 200 ms, it SHOULD issue a provisional (1xx)
   response as soon as possible. Stateless proxies MUST NOT issue
   provisional responses on their own.



Handley/Schulzrinne/Schooler/Rosenberg                       [Page 95]

Internet Draft                    SIP                      July 13, 2000


   Responses are mapped to requests by the matching To, From, Call-ID,
   CSeq headers and the branch parameter of the first Via header.
   Responses terminate request retransmissions even if they have Via
   headers that cause them to be delivered to an upstream client.

   A stateful proxy may receive a response that it does not have state
   for, that is, where it has no a record of an associated request. If
   the Via header field indicates that the upstream server used TCP, the
   proxy actively opens a TCP connection to that address. Thus, proxies
   have to be prepared to receive responses on the incoming side of
   passive TCP connections, even though most responses will arrive on
   the incoming side of an active connection. (An active connection is a
   TCP connection initiated by the proxy, a passive connection is one
   accepted by the proxy, but initiated by another entity.)

   100 responses SHOULD NOT be forwarded, other 1xx responses MAY be
   forwarded, possibly after the server eliminates responses with status
   codes that had already been sent earlier. 2xx responses are forwarded
   according to the Via header. Once a stateful proxy has received a 2xx
   response, it MUST NOT forward non-2xx final responses.  Responses
   with status 300 and higher are retransmitted by each stateful proxy
   or UAS until the next upstream proxy or UAC sends an ACK (see below
   for timing details) or CANCEL.

   A stateful proxy SHOULD maintain state for at least 32 seconds after
   the receipt of the first definitive non-200 response, in order to
   handle retransmissions of the response.


        The 32 second window is given by the maximum retransmission
        duration of 200-class responses using the default timers,
        in case the ACK is lost somewhere on the way to the called
        user agent or the next stateful proxy.

10.2 Source Addresses, Destination Addresses and Connections

10.2.1 Unicast UDP

   Responses are sent according to the rules in Section 6.47.3 or
   Section 6.47.4.


        Recall that responses are not generated by the next-hop
        stateless server, but generated by either a proxy server or
        the user agent server. Thus, the stateless proxy can only
        use the Via header field to forward the response.

10.2.2 Multicast UDP



Handley/Schulzrinne/Schooler/Rosenberg                       [Page 96]

Internet Draft                    SIP                      July 13, 2000


   Requests MAY be multicast; multicast requests likely feature a host-
   independent Request-URI. This request SHOULD be scoped to ensure it
   is not forwarded beyond the boundaries of the administrative scope.
   This MAY be done with either TTL or administrative scopes [29],
   depending on what is implemented in the network.

   A client receiving a multicast query does not have to check whether
   the host part of the Request-URI matches its own host or domain name.
   If the request was received via multicast, the response MUST be
   returned to the address listed in the maddr parameter of the Via
   header field. Generally, this will be a multicast address. Such
   multicast responses are multicast with the same TTL as the request,
   where the TTL is derived from the ttl parameter in the Via header
   (Section 6.47).

   To avoid response implosion, servers MUST NOT answer multicast
   requests with a status code other than 2xx, 401, 407, 484 or 6xx. The
   server delays its response by a random interval uniformly distributed
   between zero and one second. Servers MAY suppress responses if they
   hear a lower-numbered or 6xx response from another group member prior
   to sending. Servers do not respond to CANCEL requests received via
   multicast to avoid request implosion. A proxy or UAC SHOULD send a
   CANCEL on receiving the first 2xx, 401, 407 or 6xx response to a
   multicast request.


        Server response suppression is a MAY since it requires a
        server to violate some basic message processing rules. Lets
        say A sends a multicast request, and it is received by B,
        C, and D. B sends a 200 response. The topmost Via field in
        the response will contain the address of A. C will also
        receive this response, and could use it to suppress its own
        response. However, C would normally not examine this
        response, as the topmost Via is not its own. Normally, a
        response received with an incorrect topmost Via MUST be
        dropped, but not in this case. To distinguish this packet
        from a misrouted or multicast looped packet is fairly
        complex, and for this reason the procedure is a MAY. The
        CANCEL, instead, provides a simpler and more standard way
        to perform response suppression. It is for this reason that
        the use of CANCEL here is a SHOULD.

10.3 TCP

   A single TCP connection can serve one or more SIP transactions. A
   transaction contains zero or more provisional responses followed by
   one or more final responses. (Typically, transactions contain exactly
   one final response, but there are exceptional circumstances, where,



Handley/Schulzrinne/Schooler/Rosenberg                       [Page 97]

Internet Draft                    SIP                      July 13, 2000


   for example, multiple 200 responses can be generated.)

   The client SHOULD keep the connection open at least until the first
   final response arrives. If the client closes or resets the TCP
   connection prior to receiving the first final response, the server
   treats this action as equivalent to a CANCEL request.


        This behavior makes it less likely that malfunctioning
        clients cause a proxy server to keep connection state
        indefinitely.

   The server SHOULD NOT close the TCP connection until it has sent its
   final response, at which point it MAY close the TCP connection if it
   wishes to. However, normally it is the client's responsibility to
   close the connection.  If the server closes the connection
   prematurely, the client SHOULD interpret this as being equivalent to
   a 500 (Server Internal Error) response.

   If the server leaves the connection open, and if the client so
   desires it MAY re-use the connection for further SIP requests or for
   requests from the same family of protocols (such as HTTP or stream
   control commands).

   If a server needs to return a response to a client and no longer has
   a connection open to that client, it MAY open a connection to the
   address listed in the Via header. Thus, a proxy or user agent MUST be
   prepared to receive both requests and responses on a "passive"
   connection.

10.4 Reliability for Requests Other Than INVITE

10.4.1 UDP

   A SIP client using UDP SHOULD retransmit requests other than INVITE
   or ACK with an exponential backoff, starting at a T1 second interval,
   doubling the interval for each packet, and capping off at a T2 second
   interval. This means that after the first packet is sent, the second
   is sent T1 seconds later, the next 2*T1 seconds after that, the next
   4*T1 seconds after that, and so on, until the interval reaches T2.
   Subsequent retransmissions are spaced by T2 seconds. If the client
   receives a provisional response, it continues to retransmit the
   request, but with an interval of T2 seconds.  Retransmissions cease
   when the client has sent a total of eleven packets, or receives a
   definitive response. Default values for T1 and T2 are 500 ms and 4 s,
   respectively. Clients MAY use larger values, but SHOULD NOT use
   smaller ones. Servers retransmit the response upon receipt of a
   request retransmission. After the server sends a final response, it



Handley/Schulzrinne/Schooler/Rosenberg                       [Page 98]

Internet Draft                    SIP                      July 13, 2000


   cannot be sure the client has received the response, and thus SHOULD
   cache the results for at least 10*T2 seconds to avoid having to, for
   example, contact the user or location server again upon receiving a
   request retransmission.


        Use of the exponential backoff is for congestion control
        purposes. However, the back-off must cap off, since request
        retransmissions are used to trigger response
        retransmissions at the server. Without a cap, the loss of a
        single response could significantly increase transaction
        latencies.

   The value of the initial retransmission timer is smaller than that
   that for TCP since it is expected that network paths suitable for
   interactive communications have round-trip times smaller than 500 ms.
   For congestion control purposes, the retransmission count has to be
   bounded.  Given that most transactions are expected to consist of one
   request and a few responses, round-trip time estimation is not likely
   to be very useful. If RTT estimation is desired to more quickly
   discover a missing final response, each request retransmission needs
   to be labeled with its own Timestamp (Section 6.43), returned in the
   response. The server caches the result until it can be sure that the
   client will not retransmit the same request again.

   Each server in a proxy chain generates its own final response to a
   CANCEL request. The server responds immediately upon receipt of the
   CANCEL request rather than waiting until it has received final
   responses from the CANCEL requests it generates.

   BYE and OPTIONS final responses are generated by redirect and user
   agent servers; REGISTER final responses are generated by registrars.
   Note that in contrast to the reliability mechanism described in
   Section 10.5, responses to these requests are not retransmitted
   periodically and not acknowledged via ACK.

10.4.2 TCP

   Clients using TCP do not need to retransmit requests, but MAY give up
   after receiving no response for an extended period of time.

10.5 Reliability for INVITE Requests

   Special considerations apply for the INVITE method.

        1.   After receiving an invitation, considerable time can elapse
             before the server can determine the outcome. For example,
             if the called party is "rung" or extensive searches are



Handley/Schulzrinne/Schooler/Rosenberg                       [Page 99]

Internet Draft                    SIP                      July 13, 2000


             performed, delays between the request and a definitive
             response can reach several tens of seconds. If either
             caller or callee are automated servers not directly
             controlled by a human being, a call attempt could be
             unbounded in time.

        2.   If a telephony user interface is modeled or if we need to
             interface to the PSTN, the caller's user interface will
             provide "ringback", a signal that the callee is being
             alerted. (The status response 180 (Ringing) MAY be used to
             initiate ringback.) Once the callee picks up, the caller
             needs to know so that it can enable the voice path and stop
             ringback. The callee's response to the invitation could get
             lost. Unless the response is transmitted reliably, the
             caller will continue to hear ringback while the callee
             assumes that the call exists.

        3.   The client has to be able to terminate an on-going request,
             e.g., because it is no longer willing to wait for the
             connection or search to succeed. The server will have to
             wait several retransmission intervals to interpret the lack
             of request retransmissions as the end of a call. If the
             call succeeds shortly after the caller has given up, the
             callee will "pick up the phone" and not be "connected".

10.5.1 UDP

   For UDP, A SIP client SHOULD retransmit a SIP INVITE request with an
   interval that starts at T1 seconds, and doubles after each packet
   transmission. The client ceases retransmissions if it receives a
   provisional or definitive response, or once it has sent a total of
   seven request packets. A UAC MAY send a BYE or CANCEL request after
   the seventh retransmission. It is RECOMMENDED to send both. (This
   avoids call establishment in case the network path loses packets
   asymmetrically.)

   A server which transmits a provisional response should retransmit it
   upon reception of a duplicate request. A server which transmits a
   final response should retransmit it with an interval that starts at
   T1 seconds, and doubles for each subsequent packet until it reaches
   T2 seconds.  Response retransmissions cease when an ACK request is
   received or the response has been retransmitted seven times. The
   value of a final response is not changed by the arrival of a BYE or
   CANCEL request.

   Only the user agent client generates an ACK for 2xx final responses,
   If the response contained a Contact header field, the ACK MAY be sent
   to the address listed in that Contact header field. If the response



Handley/Schulzrinne/Schooler/Rosenberg                      [Page 100]

Internet Draft                    SIP                      July 13, 2000


   did not contain a Contact header, the client uses the same To header
   field and Request-URI as for the INVITE request and sends the ACK to
   the same destination as the original INVITE request. ACKs for final
   responses other than 2xx are sent to the same server that the
   original request was sent to, using the same Request-URI as the
   original request. Note, however, that the To header field in the ACK
   is copied from the response being acknowledged, not the request, and
   thus MAY additionally contain the tag parameter. Also note than
   unlike 2xx final responses, a proxy generates an ACK for non-2xx
   final responses.

   The ACK request MUST NOT be acknowledged to prevent a response-ACK
   feedback loop. Fig. 11 and 12 show the client and server state
   diagram for INVITE transactions. The "terminated" event occurs if the
   server receives either a CANCEL or BYE request. Note that the state
   diagram only shows the behavior for the INVITE transaction; the
   responses for BYE and CANCEL are not shown and follow the rules laid
   in Section 10.4.




        The mechanism in Sec. 10.4 would not work well for INVITE
        because of the long delays between INVITE and a final
        response. If the 200 response were to get lost, the callee
        would believe the call to exist, but the voice path would
        be dead since the caller does not know that the callee has
        picked up. Thus, the INVITE retransmission interval would
        have to be on the order of a second or two to limit the
        duration of this state confusion. Retransmitting the
        response with an exponential back-off helps ensure that the
        response is received, without placing an undue burden on
        the network.

10.5.2 TCP

   A user agent using TCP MUST NOT retransmit requests, but uses the
   same algorithm as for UDP (Section 10.5.1) to retransmit responses
   until it receives an ACK. A client MAY give up on the request if
   there is no response within a client-defined timeout interval.


        It is necessary to retransmit 2xx responses as their
        reliability is assured end-to-end only. If the chain of
        proxies has a UDP link in the middle, it could lose the
        response, with no possibility of recovery. For simplicity,
        we also retransmit non-2xx responses, although that is not
        strictly necessary.



Handley/Schulzrinne/Schooler/Rosenberg                      [Page 101]

Internet Draft                    SIP                      July 13, 2000




              +===========+
              *           *
              *  Initial  *
              *           *
              +===========+
                    |
                    |    -
                    |  INVITE
                    |
                    v
              *************
    T1*2^n <--*           *
    INVITE -->*  Calling  *--------+
              *           *        |
              *************        |
                :   |              |
  ..............:   | 1xx      xxx |
  : 7 INVITE sent   |  -       ACK |
  :                 |              |
  :                 v              |
  :           *************        |
  :           *   Call    *        |
  :           * proceeding*<->1xx  |
  :           *           *        |
  :           *************        |
  :                 |              |
  :                 |<-------------+
  :..............   |
                .   v
              *************
      xxx  <--*           *
      ACK  -->* Completed *
              *           *
              *************

 event (xxx=status)
     message


   Figure 11: State transition diagram of client for INVITE method


10.6 Reliability for ACK Requests

   The ACK request does not generate responses. It is only generated
   when a response to an INVITE request arrives (see Section 10.5). This



Handley/Schulzrinne/Schooler/Rosenberg                      [Page 102]

Internet Draft                    SIP                      July 13, 2000




                      +===============+                   
                      *               *
                      *   Initial     *
                      *               *
                      +===============+
                              !
                              ! INVITE/1xx
                              !
                              !
                              v
                      *****************
             |------->*               *<----------|
        INVITE/1xx    *  Proceeding   *    status change/1xx
             ^--------*               *-----------^
;;;;;;;;;;;;;;;;;;;;;;*****************
; terminated/487            !   !
:                           !   !
;                           !   ! 
;            failure/>=300  !   !  picks up/2xx
;             +-------------+   +-----------+
;             v                             v
;        ***********                   ***********
;INVITE/>*         *<-min(T1*2^n,T2)/->*         *<-----|
;status <* failure *->    status     <-* success *  INVITE/2xx
;        *         *                   *         *------^
;;;;;;;;;***********                   ***********
              !                             !
              !                             !
              !                             !
              !                             !
              +--------------+--------------+
event/message sent           ! ACK/-
                             v
                     *****************
                |--->*               *
              ACK/-  *   Confirmed   *
                ^--->*               *
                     *****************


   Figure 12: State transition diagram of server for INVITE method


   behavior is independent of the transport protocol. Note that the ACK
   request MAY take a different path than the original INVITE request,
   and MAY even cause a new TCP connection to be opened in order to send



Handley/Schulzrinne/Schooler/Rosenberg                      [Page 103]

Internet Draft                    SIP                      July 13, 2000


   it.

10.7 ICMP Handling

   Handling of ICMP messages in the case of UDP messages is
   straightforward. For requests, a host, network, port, or protocol
   unreachable error SHOULD be treated as if a 400-class response was
   received. For responses, these errors SHOULD cause the server to
   cease retransmitting the response.

   Source quench ICMP messages SHOULD be ignored. TTL exceeded errors
   SHOULD be ignored. Parameter problem errors SHOULD be treated as if a
   400-class response was received.

11 Behavior of SIP User Agents

   This section describes the rules for user agent client and servers
   for generating and processing requests and responses.

11.1 Caller Issues Initial INVITE Request

   When a user agent client desires to initiate a call, it formulates an
   INVITE request. The To field in the request contains the address of
   the callee, and remains unaltered as the request traverses proxies.
   The Request-URI contains the same address, but may be rewritten by
   proxies. The From field contains the address of the caller. If the
   From address can appear in requests generated by other user agent
   clients for the same call, the caller MUST insert the tag parameter
   in the From field. A UAC MAY optionally add a Contact header
   containing an address where it would like to be contacted for
   transactions from the callee back to the caller.

11.2 Callee Issues Response

   When the initial INVITE request is received at the callee, the callee
   can accept, redirect, or reject the call. In all of these cases, it
   formulates a response. The response MUST copy the To, From, Call-ID,
   CSeq and Via fields from the request. Additionally, the responding
   UAS MUST add the tag parameter to the To field in the response if the
   request contained more than one Via header field. Since a request
   from a UAC may fork and arrive at multiple hosts, the tag parameter
   serves to distinguish, at the UAC, multiple responses from different
   UAS's. The UAS MAY add a Contact header field in the response. It
   contains an address where the callee would like to be contacted for
   subsequent transactions, including the ACK for the current INVITE.
   The UAS stores the values of the To and From field, including any
   tags. These become the local and remote addresses of the call leg,
   respectively.



Handley/Schulzrinne/Schooler/Rosenberg                      [Page 104]

Internet Draft                    SIP                      July 13, 2000


11.3 Caller Receives Response to Initial Request

   Multiple responses may arrive at the UAC for a single INVITE request,
   due to a forking proxy. Each response is distinguished by the "tag"
   parameter in the To header field, and each represents a distinct call
   leg. The caller MAY choose to acknowledge or terminate the call with
   each responding UAS. To acknowledge, it sends an ACK request, and to
   terminate it sends a BYE request.  The To header field in the ACK or
   BYE MUST be the same as the To field in the 200 response, including
   any tag. The From header field MUST be the same as the From header
   field in the 200 (OK) response, including any tag. The Request-URI of
   the ACK or BYE request MAY be set to whatever address was found in
   the Contact header field in the 200 (OK) response, if present.
   Alternately, a UAC may copy the address from the To header field into
   the Request-URI. The UAC also notes the value of the To and From
   header fields in each response. For each call leg, the To header
   field becomes the remote address, and the From header field becomes
   the local address.

11.4 Caller or Callee Generate Subsequent Requests

   Once the call has been established, either the caller or callee MAY
   generate INVITE or BYE requests to change or terminate the call.
   Regardless of whether the caller or callee is generating the new
   request, the header fields in the request are set as follows. For the
   desired call leg, the To header field is set to the remote address,
   and the From header field is set to the local address (both including
   any tags). A UAC copies the tag from the final response into the ACK,
   but it MUST NOT copy the tag into any subsequent requests unless the
   response was a 200-class response to an INVITE request. The To field
   of CANCEL requests always contain exactly the same value as the
   request it is cancelling.

   The Contact header field MAY be different than the Contact header
   field sent in a previous response or request.  The Request-URI MAY be
   set to the value of the Contact header field received in a previous
   request or response from the remote party.

   The network destination and Request-URI of requests is determined
   according to the following rules:

        o If the response from the previous request contained a Record-
          Route header field, the UAC sends the request to the last
          entry in the list and removes that entry. As described in
          Section 6.35, the Request-URI is set to that value.

        o Otherwise, if the response from the previous response
          contained a Contact header field, the request is directed to



Handley/Schulzrinne/Schooler/Rosenberg                      [Page 105]

Internet Draft                    SIP                      July 13, 2000


          the host and port identified there. The Request-URI is set to
          the value of the Contact header. The request does not contain
          a Route header field in this case.

        o Otherwise, if the UAC is configured with the address of an
          outbound proxy server, the UAC sends the request there. The
          Request-URI contains the same URL as the To header field if
          the UAC is the callee. It copies the URL and To header field
          from the caller's From header if the UAC is the callee.

        o Otherwise, if the UAC is the caller, it copies the To header
          field into the Request-URI If the UAC is the callee, it copies
          the From value of the caller's request into the To header
          field and the Request-URI. In both cases, the UAC sends the
          request to the host identified in the URL.

   If a UAC does not support DNS resolution or the full Record-
   Route/Route mechanism, it MAY send all requests to a locally
   configured outbound proxy. In that case, that proxy behaves as
   described above. The UAC MUST, however, perform the mapping of
   Record-Route to Route header fields and MUST include all Route header
   fields, i.e., the UAC does not remove the first Route header field.

11.5 Receiving Subsequent Requests

   When a request is received during a call, the following checks are
   made:

        1.   If the Call-ID is new, the request is for a new call,
             regardless of the values of the To and From header fields.

             It is possible that the To header in an INVITE request has
             a tag, but the UAS believes this to be a new call. This
             will occur if the UAS crashed and rebooted in the middle of
             a call, and the UAC has sent what it believes to be a re-
             INVITE. The UAS MAY either accept or reject the request.
             Accepting the request provides robustness, so that calls
             can persist even through crashes. UAs wishing to support
             this capability must choose monotonically increasing CSeq
             numbers even across reboots. This is because subsequent
             requests from the crashed-and-rebooted UA towards the other
             UA need to have a CSeq number higher than previous requests
             in that direction.

             Note also that the crashed-and-rebooted UA will have lost
             any Route headers which would need to be inserted into a
             subsequent request. Therefore, it is possible that the
             requests may not be properly forwarded by proxies.



Handley/Schulzrinne/Schooler/Rosenberg                      [Page 106]

Internet Draft                    SIP                      July 13, 2000


             RTP media agents allowing restarts need to be robust
             by accepting out-of-range timestamps and sequence
             numbers.

        2.   If the Call-ID exists, the request is for an existing call.
             If the To, From, Call-ID, and CSeq values exactly match
             (including tags) those of any requests received previously,
             the request is a retransmission.

        3.   If there was no match to the previous step, the To and From
             fields are compared against existing call leg local and
             remote addresses. If there is a match, and the CSeq in the
             request is higher than the last CSeq received on that leg,
             the request is a new transaction for an existing call leg.

12 Behavior of SIP Proxy and Redirect Servers

   This section describes behavior of SIP redirect and proxy servers in
   detail. Proxy servers can "fork" connections, i.e., a single incoming
   request spawns several outgoing (client) requests.

12.1 Redirect Server

   A redirect server does not issue any SIP requests of its own. After
   receiving a request other than CANCEL, the server gathers the list of
   alternative locations and returns a final response of class 3xx or it
   refuses the request. For well-formed CANCEL requests, it SHOULD
   return a 2xx response. This response ends the SIP transaction. The
   redirect server maintains transaction state for the whole SIP
   transaction. It is up to the client to detect forwarding loops
   between redirect servers.

12.2 User Agent Server

   User agent servers behave similarly to redirect servers, except that
   they also accept requests and can return a response of class 2xx.

12.3 Proxy Server

   This section outlines processing rules for proxy servers. A proxy
   server can either be stateful or stateless. When stateful, a proxy
   remembers the incoming request which generated outgoing requests, and
   the outgoing requests. A stateless proxy forgets all information once
   an outgoing request is generated. A forking proxy SHOULD be stateful.
   Proxies that accept TCP connections MUST be stateful when handling
   the TCP connection.





Handley/Schulzrinne/Schooler/Rosenberg                      [Page 107]

Internet Draft                    SIP                      July 13, 2000


        Otherwise, if the proxy were to lose a request, the TCP
        client would never retransmit it.

   A stateful proxy SHOULD NOT become stateless until after it sends a
   definitive response upstream, and at least 32 seconds after it
   received a definitive response.

   A stateful proxy acts similar to a virtual UAS/UAC, but cannot be
   viewed as just a UAS and UAC glued together at the back. (In
   particular, it does not originate requests except ACK and CANCEL.)
   It implements the server state machine when receiving requests, and
   the client state machine for generating outgoing requests, with the
   exception of receiving a 2xx response to an INVITE. Instead of
   generating an ACK, the 2xx response is always forwarded upstream
   towards the caller. Furthermore, ACK's for 200 responses to INVITE's
   are always proxied downstream towards the UAS, as they would be for a
   stateless proxy.

   A stateless proxy forwards every request it receives downstream, and
   every response it receives upstream.

12.3.1 Proxying Requests

   A proxy server MUST check for forwarding loops before proxying a
   request.  A request has been looped if the server finds its own
   address in the Via header field and the hash computation over the
   fields enumerated in Section 6.47.5 yields the same value as the hash
   part of the "branch" parameter in the Via entry containing the proxy
   server's address.

   A proxy server MUST NOT forward a request to a multicast group which
   already appears in any of the Via headers.

   The To, From, Call-ID, and Contact tags are copied exactly from the
   original request. The proxy SHOULD change the Request-URI to indicate
   the server where it intends to send the request.

   A proxy server always inserts a Via header field containing its own
   address into those requests that are caused by an incoming request.
   Each proxy MUST insert a "branch" parameter (Section 6.47).

12.3.2 Proxying Responses

   A proxy only processes a response if the topmost Via field matches
   one of its addresses. A response with a non-matching top Via field
   MUST be dropped.

12.3.3 Stateless Proxy: Proxying Responses



Handley/Schulzrinne/Schooler/Rosenberg                      [Page 108]

Internet Draft                    SIP                      July 13, 2000


   A stateless proxy removes its own Via field, and checks the address
   in the next Via field. In the case of UDP, the response is sent to
   the address listed in the "maddr" tag if present, otherwise to the
   "received" tag if present, and finally to the address in the "sent-
   by" field. A proxy MUST remain stateful when handling requests
   received via TCP.

   A stateless proxy MUST NOT generate its own provisional responses.

12.3.4 Stateful Proxy: Receiving Requests

   When a stateful proxy receives a request, it checks the To, From
   (including tags), Call-ID and CSeq against existing request records.
   If the tuple exists, the request is a retransmission. The provisional
   or final response sent previously is retransmitted, as per the server
   state machine. If the tuple does not exist, the request corresponds
   to a new transaction, and the request should be proxied.

   A stateful proxy server MAY generate its own provisional (1xx)
   responses.

12.3.5 Stateful Proxy: Receiving ACKs

   When an ACK request is received, it is either processed locally or
   proxied. To make this determination, the To, From, CSeq and Call-ID
   fields are compared against those in previous requests. If there is
   no match, the ACK request is proxied as if it were an INVITE request.
   If there is a match, and if the server had ever sent a 200 response
   upstream, the ACK is proxied.  If the server had never sent any
   responses upstream, the ACK is also proxied. If the server had sent a
   3xx, 4xx, 5xx or 6xx response, but no 2xx response, the ACK is
   processed locally if the tag in the To field of the ACK matches the
   tag sent by the proxy in the response.

12.3.6 Stateful Proxy: Receiving Responses

   When a proxy server receives a response that has passed the Via
   checks, the proxy server checks the To (without the tag), From
   (including the tag), Call-ID and CSeq against values seen in previous
   requests. If there is no match, the response is forwarded upstream to
   the address listed in the Via field. If there is a match, the
   "branch" tag in the Via field is examined. If it matches a known
   branch identifier, the response is for the given branch, and
   processed by the virtual client for the given branch. Otherwise, the
   response is dropped.

   A stateful proxy should obey the rules in Section 12.4 to determine
   if the response should be proxied upstream. If it is to be proxied,



Handley/Schulzrinne/Schooler/Rosenberg                      [Page 109]

Internet Draft                    SIP                      July 13, 2000


   the same rules for stateless proxies above are followed, with the
   following addition for TCP. If a request was received via TCP
   (indicated by the protocol in the top Via header), the proxy checks
   to see if it has a connection currently open to that address. If so,
   the response is sent on that connection.  Otherwise, a new TCP
   connection is opened to the address and port in the Via field, and
   the response is sent there. Note that this implies that a UAC or
   proxy MUST be prepared to receive responses on the incoming side of a
   TCP connection. Definitive non 200-class responses MUST be
   retransmitted by the proxy, even over a TCP connection.

12.3.7 Stateless, Non-Forking Proxy

   Proxies in this category issue at most a single unicast request for
   each incoming SIP request, that is, they do not "fork" requests.
   However, servers MAY choose to always operate in a mode that allows
   issuing of several requests, as described in Section 12.4.

   The server can forward the request and any responses. It does not
   have to maintain any state for the SIP transaction. Reliability is
   assured by the next redirect or stateful proxy server in the server
   chain.

   A proxy server SHOULD cache the result of any address translations
   and the response to speed forwarding of retransmissions. After the
   cache entry has been expired, the server cannot tell whether an
   incoming request is actually a retransmission of an older request.
   The server will treat it as a new request and commence another
   search.

12.4 Forking Proxy

   The server must respond to the request (other than ACK) immediately
   with a 100 (Trying) response if it expects to take more than 200 ms
   to obtain a final response.

   Successful responses to an INVITE request MAY contain a Contact
   header field so that the following ACK or BYE bypasses the proxy
   search mechanism. If the proxy requires future requests to be routed
   through it, it adds a Record-Route header to the request (Section
   6.35).

   The following C-code describes the behavior of a proxy server issuing
   several requests in response to an incoming INVITE request with
   method R which is to be proxied to a list of N destination enumerated
   in ' address .I expires

   The function request(r, a, b) sends a SIP request of type r to



Handley/Schulzrinne/Schooler/Rosenberg                      [Page 110]

Internet Draft                    SIP                      July 13, 2000


   address a, with branch id b. await_response() waits until a response
   is received and returns the response. close(a) closes the TCP
   connection to client with address a. response(r) sends a response to
   the client. ismulticast() returns 1 if the location is a multicast
   address and zero otherwise.  The variable timeleft indicates the
   amount of time left until the maximum response time has expired. The
   variable recurse indicates whether the server will recursively try
   addresses returned through a 3xx response. A server MAY decide to
   recursively try only certain addresses, e.g., those which are within
   the same domain as the proxy server. Thus, an initial multicast
   request can trigger additional unicast requests.


     /* request type */
     typedef enum {INVITE, ACK, BYE, OPTIONS, CANCEL, REGISTER} Method;

     process_request(Method R, int N, address_t address[], int expires)
     {
       struct {
         char *branch;         /* branch token */
         int branch_seq;       /* branch sequence number part */
         int done;             /* has responded */
       } outgoing[];
       char *location[];       /* list of locations */
       int heard = 0;          /* number of sites heard from */
       int class;              /* class of status code */
       int timeleft = expires; /* expiration value */
       int loc = 0;            /* number of locations */
       struct {                /* response */
         int status;           /* response: CANCEL=-1 */
         int locations;        /* number of redirect locations */
         char *location[];     /* redirect locations */
         address_t a;          /* address of respondent */
         char *branch;         /* branch token */
         int branch_seq;       /* branch sequence number */
       } r, best;              /* response, best response */
       int i;

       best.status = 1000;
       for (i = 0; i < N; i++) {
         request(R, address[i], i);
         outgoing[i].done = 0;
         outgoing[i].branch = "";
         outgoing[i].branch_seq = i;
       }

       while (timeleft > 0 && heard < N) {
         r = await_response();



Handley/Schulzrinne/Schooler/Rosenberg                      [Page 111]

Internet Draft                    SIP                      July 13, 2000


         class = r.status / 100;

         /* If final response, mark branch as done. */
         if (class >= 2) {
           heard++;
           for (i = 0; i < N; i++) {
             if (r.branch_seq == outgoing[i].branch_seq) {
               outgoing[i].done = 1;
               break;
             }
           }
         }
         /* CANCEL: respond, fork and wait for responses */
         /* terminate INVITE with 40
         else if (class < 0) {
           best.status = 200;
           response(best);
           for (i = 0; i < N; i++) {
             if (!outgoing[i].done)
               request(CANCEL, address[i], outgoing[i].branch);
           }
           best.status = -1;
         }

         /* Send an ACK */
         if (class != 2) {
           if (R == INVITE) request(ACK, r.a, r.branch);
         }

         if (class == 2) {
           if (r.status < best.status) best = r;
           break;
         }
         else if (class == 3) {
           /* A server MAY optionally recurse.  The server MUST check
            * whether it has tried this location before and whether the
            * location is part of the Via path of the incoming request.
            * This check is omitted here for brevity.  Multicast locations
            * MUST NOT be returned to the client if the server is not
            * recursing.
            */
           if (recurse) {
             multicast = 0;
             N += r.locations;
             for (i = 0; i < r.locations; i++) {
               request(R, r.location[i]);
             }
           } else if (!ismulticast(r.location)) {



Handley/Schulzrinne/Schooler/Rosenberg                      [Page 112]

Internet Draft                    SIP                      July 13, 2000


             best = r;
           }
         }
         else if (class == 4) {
           if (best.status >= 400) best = r;
         }
         else if (class == 5) {
           if (best.status >= 500) best = r;
         }
         else if (class == 6) {
           best = r;
           break;
         }
       }

       /* We haven't heard anything useful from anybody. */
       if (best.status == 1000) {
         best.status = 408; /* request expired */
       }
       if (best.status/100 != 3) loc = 0;
       response(best);
     }



   Responses are processed as follows. The process completes (and state
   can be freed) when all requests have been answered by final status
   responses (for unicast) or 60 seconds have elapsed (for multicast). A
   proxy MAY send a CANCEL to all incomplete branches and return the
   best available final status to the client if not all responses have
   been received after 60 seconds or the expiration period specified in
   the Expires header field of the request. If no responses have been
   received, the proxy returns a 408 (Timeout) response to the client.

   When forwarding responses, a proxy MUST forward the whole response,
   including all header fields of the selected response as well as the
   body.

        1xx: The proxy SHOULD forward provisional responses greater than
             100 upstream towards the client and SHOULD NOT forward 100
             (Trying) responses.

        2xx: The proxy MUST forward the response upstream towards the
             client, without sending an ACK downstream. After receiving
             a 2xx, the server MAY terminate all other pending requests
             by sending a CANCEL request and closing the TCP connection,
             if applicable.  (Terminating pending requests is advisable
             as searches consume resources. Also, INVITE requests could



Handley/Schulzrinne/Schooler/Rosenberg                      [Page 113]

Internet Draft                    SIP                      July 13, 2000


             "ring" on a number of workstations if the callee is
             currently logged in more than once.)

        3xx: For INVITE requests, the proxy MUST send an ACK.  It MAY
             recurse on the listed Contact addresses. Otherwise, the
             lowest-numbered response is returned if there were no 2xx
             or 6xx responses.

             Location lists are not merged as that would prevent
             forwarding of authenticated responses. Also, responses
             can have message bodies, so that merging is not
             feasible.

        4xx, 5xx: For INVITE requests, the proxy MUST send an ACK. It
             remembers the response if it has a lower status code class
             than any previous 4xx and 5xx response. On completion, a
             response with the lowest response class is returned if
             there were no 2xx, 3xx or 6xx responses. Within the set of
             responses from the lowest-numbered class, the proxy server
             may choose any response.

             The proxy SHOULD collect all WWW-Authenticate and Proxy-
             Authenticate headers from all 401 and 407 responses and
             return all of them in the response if either 401 or 407 is
             the lowest-numbered response.

        6xx: For INVITE requests, the proxy sends an ACK. It forwards
             the 6xx response unless a 2xx response has been received.
             Other pending requests MAY be terminated with CANCEL as
             described for 2xx responses. Unlike for 2xx responses, only
             one 6xx response is forwarded, since ACKs are generated
             locally.

   A proxy server forwards any response for Call-IDs for which it does
   not have a pending transaction according to the response's Via
   header. User agent servers respond to BYE requests for unknown call
   legs with status code 481 (Transaction Does Not Exist); they drop ACK
   requests with unknown call legs silently.

   Special considerations apply for choosing forwarding destinations for
   ACK and BYE requests. In most cases, these requests will bypass
   proxies and reach the desired party directly, keeping proxies from
   having to make forwarding decisions.

   A proxy MAY maintain call state for a period of its choosing. If a
   proxy still has list of destinations that it forwarded the last
   INVITE to, it SHOULD direct ACK requests only to those downstream
   servers.



Handley/Schulzrinne/Schooler/Rosenberg                      [Page 114]

Internet Draft                    SIP                      July 13, 2000


13 Security Considerations

13.1 Confidentiality and Privacy: Encryption

13.1.1 End-to-End Encryption

   SIP requests and responses can contain sensitive information about
   the communication patterns and communication content of individuals.
   The SIP message body MAY also contain encryption keys for the session
   itself. SIP supports three complementary forms of encryption to
   protect privacy:

        o End-to-end encryption of the SIP message body and certain
          sensitive header fields;

        o hop-by-hop encryption to prevent eavesdropping that tracks who
          is calling whom;

        o hop-by-hop encryption of Via fields to hide the route a
          request has taken.

   Not all of the SIP request or response can be encrypted end-to-end
   because header fields such as To and Via need to be visible to
   proxies so that the SIP request can be routed correctly.  Hop-by-hop
   encryption encrypts the entire SIP request or response on the wire so
   that packet sniffers or other eavesdroppers cannot see who is calling
   whom. Hop-by-hop encryption can also encrypt requests and responses
   that have been end-to-end encrypted. Note that proxies can still see
   who is calling whom, and this information is also deducible by
   performing a network traffic analysis, so this provides a very
   limited but still worthwhile degree of protection.

   SIP Via fields are used to route a response back along the path taken
   by the request and to prevent infinite request loops. However, the
   information given by them can also provide useful information to an
   attacker. Section 6.26 describes how a sender can request that Via
   fields be encrypted by cooperating proxies without compromising the
   purpose of the Via field.

   End-to-end encryption relies on keys shared by the two user agents
   involved in the request. Typically, the message is sent encrypted
   with the public key of the recipient, so that only that recipient can
   read the message. All implementations SHOULD support PGP-based
   encryption [42] and MAY implement other schemes.

   A SIP request (or response) is end-to-end encrypted by splitting the
   message to be sent into a part to be encrypted and a short header
   that will remain in the clear. Some parts of the SIP message, namely



Handley/Schulzrinne/Schooler/Rosenberg                      [Page 115]

Internet Draft                    SIP                      July 13, 2000


   the request line, the response line and certain header fields marked
   with "n" in the "enc." column in Table 4 and 5 need to be read and
   returned by proxies and thus MUST NOT be encrypted end-to-end.
   Possibly sensitive information that needs to be made available as
   plaintext include destination address (To) and the forwarding path
   (Via) of the call. The Authorization header field MUST remain in the
   clear if it contains a digital signature as the signature is
   generated after encryption, but MAY be encrypted if it contains
   "basic" or "digest" authentication.

   Other header fields MAY be encrypted or MAY travel in the clear as
   desired by the sender. The Subject, Allow and Content-Type header
   fields will typically be encrypted. The Accept, Accept-Language,
   Date, Expires, Priority, Require, Call-ID, Cseq, and Timestamp header
   fields will remain in the clear.

   All fields that will remain in the clear MUST precede those that will
   be encrypted. The message is encrypted starting with the first
   character of the first header field that will be encrypted and
   continuing through to the end of the message body. If no header
   fields are to be encrypted, encrypting starts with the second CRLF
   pair after the last header field, as shown below. Carriage return and
   line feed characters have been made visible as "$", and the encrypted
   part of the message is outlined.


     INVITE sip:watson@boston.bell-telephone.com SIP/2.0$
     Via: SIP/2.0/UDP 169.130.12.5$
     To: T. A. Watson $
     From: A. Bell $
     Encryption: PGP version=5.0$
     Content-Length: 224$
     Call-ID: 187602141351@worcester.bell-telephone.com$
     Content-Type: message/sip
     CSeq: 488$
     $
   *******************************************************
   * Subject: Mr. Watson, come here.$                    *
   * Content-Type: application/sdp$                      *
   * $                                                   *
   * v=0$                                                *
   * o=bell 53655765 2353687637 IN IP4 128.3.4.5$        *
   * s=Mr. Watson, come here.$                           *
   * t=0 0$                                              *
   * c=IN IP4 135.180.144.94$                            *
   * m=audio 3456 RTP/AVP 0 3 4 5$                       *
   *******************************************************




Handley/Schulzrinne/Schooler/Rosenberg                      [Page 116]

Internet Draft                    SIP                      July 13, 2000


   An Encryption header field MUST be added to indicate the encryption
   mechanism used. A Content-Length field is added that indicates the
   length of the encrypted body. The encrypted body is preceded by a
   blank line as a normal SIP message body would be.

   Upon receipt by the called user agent possessing the correct
   decryption key, the message body as indicated by the Content-Length
   field is decrypted, and the now-decrypted body is appended to the
   clear-text header fields. There is no need for an additional
   Content-Length header field within the encrypted body because the
   length of the actual message body is unambiguous after decryption.

   A Content-Type indication of "message/sip" MAY be added, but will be
   overridden after receipt.

   Had no SIP header fields required encryption, the message would have
   been as below. Note that the encrypted body MUST then include a blank
   line (start with CRLF) to disambiguate between any possible SIP
   header fields that might have been present and the SIP message body.


     INVITE sip:watson@boston.bell-telephone.com SIP/2.0$
     Via: SIP/2.0/UDP 169.130.12.5$
     To: T. A. Watson $
     From: A. Bell $
     Encryption: PGP version=5.0$
     Content-Type: application/sdp$
     Content-Length: 107$
     Call-ID: 187602141351@worcester.bell-telephone.com$
     CSeq: 488$
     $
   *************************************************
   * $                                             *
   * v=0$                                          *
   * o=bell 53655765 2353687637 IN IP4 128.3.4.5$  *
   * c=IN IP4 135.180.144.94$                      *
   * m=audio 3456 RTP/AVP 0 3 4 5$                 *
   *************************************************



13.1.2 Privacy of SIP Responses

   SIP requests can be sent securely using end-to-end encryption and
   authentication to a called user agent that sends an insecure
   response.  This is allowed by the SIP security model, but is not a
   good idea.  However, unless the correct behavior is explicit, it
   would not always be possible for the called user agent to infer what



Handley/Schulzrinne/Schooler/Rosenberg                      [Page 117]

Internet Draft                    SIP                      July 13, 2000


   a reasonable behavior was. Thus, when end-to-end encryption is used
   by the request originator, the encryption key to be used for the
   response SHOULD be specified in the request (Section 6.37). If this
   were not done, it might be possible for the called user agent to
   incorrectly infer an appropriate key to use in the response. Thus, to
   prevent key-guessing becoming an acceptable strategy, we specify that
   a called user agent receiving a request that does not specify a key
   to be used for the response SHOULD send that response unencrypted.

   Any SIP header fields that were encrypted in a request SHOULD also be
   encrypted in an encrypted response. Contact response fields MAY be
   encrypted if the information they contain is sensitive, or MAY be
   left in the clear to permit proxies more scope for localized
   searches.

13.1.3 Encryption by Proxies

   Normally, proxies are not allowed to alter end-to-end header fields
   and message bodies. Proxies MAY, however, encrypt an unsigned request
   or response with the key of the call recipient.


        Proxies need to encrypt a SIP request if the end system
        cannot perform encryption or to enforce organizational
        security policies.

13.1.4 Hop-by-Hop Encryption

   SIP requests and responses MAY also be protected by security
   mechanisms at the transport or network layer. No particular mechanism
   is defined or recommended here. Two possibilities are IPSEC [43] or
   TLS [24]. The use of a particular mechanism will generally need to be
   specified out of band, through manual configuration, for example.

13.1.5 Via field encryption

   When Via header fields are to be hidden, a proxy that receives a
   request containing an appropriate "Hide: hop" header field (as
   specified in section 6.26) SHOULD encrypt the header field. As only
   the proxy that encrypts the field will decrypt it, the algorithm
   chosen is entirely up to the proxy implementor. Two methods satisfy
   these requirements:

        o The server keeps a cache of Via header fields and the
          associated To header field, and replaces the Via header field
          with an index into the cache. On the reverse path, take the
          Via header field from the cache rather than the message.




Handley/Schulzrinne/Schooler/Rosenberg                      [Page 118]

Internet Draft                    SIP                      July 13, 2000


          This is insufficient to prevent message looping, and so an
          additional ID MUST be added so that the proxy can detect
          loops. This SHOULD NOT normally be the address of the proxy as
          the goal is to hide the route, so instead a sufficiently large
          random number SHOULD be used by the proxy and maintained in
          the cache.

          It is possible for replies to get directed to the wrong
          originator if the cache entry gets reused, so great care needs
          to be taken to ensure this does not happen.

        o The server MAY use a secret key to encrypt the Via field, a
          timestamp and an appropriate checksum in any such message with
          the same secret key. The checksum is needed to detect whether
          successful decoding has occurred, and the timestamp is
          required to prevent possible replay attacks and to ensure that
          no two requests from the same previous hop have the same
          encrypted Via field.  This is the preferred solution.

13.2 Message Integrity and Access Control: Authentication

   Protective measures need to be taken to prevent an active attacker
   from modifying and replaying SIP requests and responses. The same
   cryptographic measures that are used to ensure the authenticity of
   the SIP message also serve to authenticate the originator of the
   message.  However, the "basic" and "digest" authentication mechanism
   offer authentication only, without message integrity.

   Transport-layer or network-layer authentication MAY be used for hop-
   by-hop authentication. SIP also extends the HTTP WWW-Authenticate
   (Section 6.49) and Authorization (Section 6.12) header field and
   their Proxy counterparts to include cryptographically strong
   signatures. SIP also supports the HTTP "basic" and "digest" schemes
   (see Section 14) and other HTTP authentication schemes to be defined
   that offer a rudimentary mechanism of ascertaining the identity of
   the caller.

   SIP requests MAY be authenticated using the Authorization header
   field to include a digital signature of certain header fields, the
   request method and version number and the payload, none of which are
   modified between client and called user agent. The Authorization
   header field is used in requests to authenticate the request
   originator end-to-end to proxies and the called user agent, and in
   responses to authenticate the called user agent or proxies returning
   their own failure codes. If required, hop-by-hop authentication can
   be provided, for example, by the IPSEC Authentication Header.

   SIP does not dictate which digital signature scheme is used for



Handley/Schulzrinne/Schooler/Rosenberg                      [Page 119]

Internet Draft                    SIP                      July 13, 2000


   authentication, but does define how to provide authentication using
   PGP in Section 15. As indicated above, SIP implementations MAY also
   use "basic" and "digest" authentication and other authentication
   mechanisms defined for HTTP [44]. Note that "basic" authentication
   has severe security limitations. The following does not apply to
   these schemes.

   To cryptographically sign a SIP request, the order of the SIP header
   fields is important. When an Authorization header field is present,
   it indicates that all header fields following the Authorization
   header field have been included in the signature.  Therefore, hop-
   by-hop header fields which MUST or SHOULD be modified by proxies MUST
   precede the Authorization header field as they will generally be
   modified or added-to by proxy servers.  Hop-by-hop header fields
   which MAY be modified by a proxy MAY appear before or after the
   Authorization header. When they appear before, they MAY be modified
   by a proxy. When they appear after, they MUST NOT be modified by a
   proxy. To sign a request, a client constructs a message from the
   request method (in upper case) followed, without LWS, by the SIP
   version number, followed, again without LWS, by the request headers
   to be signed and the message body.  The message thus constructed is
   then signed.

   For example, if the SIP request is to be:

   INVITE sip:watson@boston.bell-telephone.com SIP/2.0
   Via: SIP/2.0/UDP 169.130.12.5
   Authorization: PGP version=5.0, signature=...
   From: A. Bell 
   To: T. A. Watson 
   Call-ID: 187602141351@worcester.bell-telephone.com
   Subject: Mr. Watson, come here.
   Content-Type: application/sdp
   Content-Length: ...

   v=0
   o=bell 53655765 2353687637 IN IP4 128.3.4.5
   s=Mr. Watson, come here.
   t=0 0
   c=IN IP4 135.180.144.94
   m=audio 3456 RTP/AVP 0 3 4 5



   Then the data block that is signed is:

   INVITESIP/2.0From: A. Bell 
   To: T. A. Watson 



Handley/Schulzrinne/Schooler/Rosenberg                      [Page 120]

Internet Draft                    SIP                      July 13, 2000


   Call-ID: 187602141351@worcester.bell-telephone.com
   Subject: Mr. Watson, come here.
   Content-Type: application/sdp
   Content-Length: ...

   v=0
   o=bell 53655765 2353687637 IN IP4 128.3.4.5
   s=Mr. Watson, come here.
   t=0 0
   c=IN IP4 135.180.144.94
   m=audio 3456 RTP/AVP 0 3 4 5



   Clients wishing to authenticate requests MUST construct the portion
   of the message below the Authorization header using a canonical form.
   This allows a proxy to parse the message, take it apart, and
   reconstruct it, without causing an authentication failure due to
   extra white space, for example. Canonical form consists of the
   following rules:

        o No short form header fields;

        o Header field names are capitalized as shown in this document;

        o No white space between the header name and the colon;

        o A single space after the colon;

        o Line termination with a CRLF;

        o No line folding;

        o No comma separated lists of header values; each must appear as
          a separate header;

        o Only a single SP between tokens, between tokens and quoted
          strings, and between quoted strings; no SP after last token or
          quoted string;

        o No LWS between tokens and separators, except as described
          above for after the colon in header fields;

        o The To and From header fields always include the < and >
          delimiters even if the display-name is empty.

   Note that if a message is encrypted and authenticated using a digital
   signature, when the message is generated encryption is performed



Handley/Schulzrinne/Schooler/Rosenberg                      [Page 121]

Internet Draft                    SIP                      July 13, 2000


   before the digital signature is generated. On receipt, the digital
   signature is checked before decryption.

   A client MAY require that a server sign its response by including a
   Require: signed-response request header field. The client indicates
   the desired authentication method via the WWW-Authenticate header.

   The correct behavior in handling unauthenticated responses to a
   request that requires authenticated responses is described in section
   13.2.1.

13.2.1 Trusting responses

   There is the possibility that an eavesdropper listens to requests and
   then injects unauthenticated responses that terminate, redirect or
   otherwise interfere with a call. (Even encrypted requests contain
   enough information to fake a response.)

   Clients need to be particularly careful with 3xx redirection
   responses.  Thus a client receiving, for example, a 301 (Moved
   Permanently) which was not authenticated when the public key of the
   called user agent is known to the client, and authentication was
   requested in the request SHOULD be treated as suspicious. The correct
   behavior in such a case would be for the called-user to form a dated
   response containing the Contact field to be used, to sign it, and
   give this signed stub response to the proxy that will provide the
   redirection. Thus the response can be authenticated correctly. A
   client SHOULD NOT automatically redirect such a request to the new
   location without alerting the user to the authentication failure
   before doing so.

   Another problem might be responses such as 6xx failure responses
   which would simply terminate a search, or "4xx" and "5xx" response
   failures.

   If TCP is being used, a proxy SHOULD treat 4xx and 5xx responses as
   valid, as they will not terminate a search. However, fake 6xx
   responses from a rogue proxy terminate a search incorrectly. 6xx
   responses SHOULD be authenticated if requested by the client, and
   failure to do so SHOULD cause such a client to ignore the 6xx
   response and continue a search.

   With UDP, the same problem with 6xx responses exists, but also an
   active eavesdropper can generate 4xx and 5xx responses that might
   cause a proxy or client to believe a failure occurred when in fact it
   did not. Typically 4xx and 5xx responses will not be signed by the
   called user agent, and so there is no simple way to detect these
   rogue responses. This problem is best prevented by using hop-by-hop



Handley/Schulzrinne/Schooler/Rosenberg                      [Page 122]

Internet Draft                    SIP                      July 13, 2000


   encryption of the SIP request, which removes any additional problems
   that UDP might have over TCP.

   These attacks are prevented by having the client require response
   authentication and dropping unauthenticated responses. A server user
   agent that cannot perform response authentication responds using the
   normal Require response of 420 (Bad Extension).

13.3 Callee Privacy

   User location and SIP-initiated calls can violate a callee's privacy.
   An implementation SHOULD be able to restrict, on a per-user basis,
   what kind of location and availability information is given out to
   certain classes of callers.

13.4 Known Security Problems

   With either TCP or UDP, a denial of service attack exists by a rogue
   proxy sending 6xx responses. Although a client SHOULD choose to
   ignore such responses if it requested authentication, a proxy cannot
   do so. It is obliged to forward the 6xx response back to the client.
   The client can then ignore the response, but if it repeats the
   request it will probably reach the same rogue proxy again, and the
   process will repeat.

14 SIP Authentication using HTTP Basic and Digest Schemes

   SIP implementations MAY use HTTP's basic and digest authentication
   mechanisms (RFC 2617 [44]) to provide a rudimentary form of security.
   This section overviews usage of these mechanisms in SIP. The basic
   operation is almost completely identical to that for HTTP [44]. This
   section outlines this operation, pointing to RFC 2617 [44] for
   details, and noting the differences when used in SIP.

14.1 Framework

   The framework for SIP authentication parallels that for HTTP (RFC
   2617 [44]). In particular, the BNF for auth-scheme, auth-param,
   challenge, realm, realm-value, and credentials is identical. The 401
   response is used by user agent servers in SIP to challenge the
   authorization of a user agent client. Additionally, registrars and
   redirect servers MAY make use of 401 responses for authorization, but
   proxies MUST NOT, and instead MAY use the 407 response. The
   requirements for inclusion of the Proxy-Authenticate, Proxy-
   Authorization, WWW-Authenticate, and Authorization in the various
   messages is identical to RFC 2617 [44].

   Since SIP does not have the concept of a canonical root URL, the



Handley/Schulzrinne/Schooler/Rosenberg                      [Page 123]

Internet Draft                    SIP                      July 13, 2000


   notion of protections spaces are interpreted differently for SIP. The
   realm is a protection domain for all SIP URIs with the same value for
   the userinfo, host and port part of the SIP Request-URI. For example:


      INVITE sip:alice.wonderland@example.com SIP/2.0
      WWW-Authenticate:  Basic realm="business"



   and


      INVITE sip:aw@example.com SIP/2.0
      WWW-Authenticate: Basic realm="business"



   define different protection realms according to this rule.

   When a UAC resubmits a request with its credentials after receiving a
   401 or 407 response, it MUST increment the CSeq header field as it
   would normally do when sending an updated request.

14.2 Basic Authentication

   The rules for basic authentication follow those defined in [44] but
   with the words "origin server" replaced with "user agent server,
   redirect server , or registrar".

   Since SIP URIs are not hierarchical, the paragraph in [44] that
   states that "all paths at or deeper than the depth of the last
   symbolic element in the path field of the Request-URI also are within
   the protection space specified by the Basic realm value of the
   current challenge" does not apply for SIP. SIP clients MAY
   preemptively send the corresponding Authorization header with
   requests for SIP URIs within the same protection realm (as defined
   above) without receipt of another challenge from the server.

14.3 Digest Authentication

   The rules for digest authentication follow those defined in [44],
   with "HTTP 1.1" replaced by "SIP/2.0" in addition to the following
   differences:

        1.   The URI included in the challenge has the following BNF:





Handley/Schulzrinne/Schooler/Rosenberg                      [Page 124]

Internet Draft                    SIP                      July 13, 2000


             URI  =  SIP-URL


        2.   The BNF in RFC 2617 has an error in that the URI is not
             enclosed in quotation marks. (The example in Section 3.5 is
             correct.) For SIP, the URI MUST be enclosed in quotation
             marks.

        3.   The BNF for digest-uri-value is:


             digest-uri-value  =  Request-URI ; as defined in Section
             4.3


        4.   The example procedure for choosing a nonce based on Etag
             does not work for SIP.

        5.   The Authentication-Info and Proxy-Authentication-Info
             fields are not used in SIP.

        6.   The text in RFC 2617 [44] regarding cache operation does
             not apply to SIP.

        7.   RFC 2617 [44] requires that a server check that the URI in
             the request line, and the URI included in the Authorization
             header, point to the same resource. In a SIP context, these
             two URI's may actually refer to different users, due to
             forwarding at some proxy.  Therefore, in SIP, a server MAY
             check that the request-uri in the Authorization header
             corresponds to a user that the server is willing to accept
             forwarded or direct calls for.

14.4 Proxy-Authentication

   The use of the Proxy-Authentication and Proxy-Authorization parallel
   that as described in [44], with one difference. Proxies MUST NOT add
   the Proxy-Authorization header. 407 responses MUST be forwarded
   upstream towards the client following the procedures for any other
   response. It is the client's responsibility to add the Proxy-
   Authorization header containing credentials for the proxy which has
   asked for authentication.


        If a proxy were to resubmit a request with a Proxy-
        Authorization header field, it would need to increment the
        CSeq in the new request. However, this would mean that the
        UAC which submitted the original request would discard a



Handley/Schulzrinne/Schooler/Rosenberg                      [Page 125]

Internet Draft                    SIP                      July 13, 2000


        response from the UAS, as the CSeq value would be
        different.

   See sections 6.32 and 6.33 for additional information on usage of
   these fields as they apply to SIP.

   It is also possible that a 401 (Unauthorized) response contains
   several challenges, from a mixture of proxies and user agent servers,
   if the request was forked.

15 SIP Security Using PGP

15.1 PGP Authentication Scheme

   The "pgp" authentication scheme is based on the model that the client
   authenticates itself with a request signed with the client's private
   key. The server can then ascertain the origin of the request if it
   has access to the public key, preferably signed by a trusted third
   party.  Implementations supporting this scheme MUST implement the
   definitions and default algorithms of RFC 2440 [45] and MAY implement
   the older version, based upon PGP 2.6, described in RFC 1991 [42].

15.1.1 The WWW-Authenticate Response Header



        WWW-Authenticate  =  "WWW-Authenticate" ":" "pgp" pgp-challenge
        pgp-challenge     =  # pgp-params
        pgp-params        =  realm | pgp-version | pgp-micalgorithm
                         |   pgp-pubalgorithm | nonce
        realm             =  "realm" "=" realm-value
        realm-value       =  quoted-string
        pgp-version       =  "version" "="
                             <"> digit *( "." digit ) *letter <">
        pgp-micalgorithm  =  "algorithm" "=" ( "md5" | "sha1" | token
                             | "ripemd160" | "MD2" | "TIGER192" | "HAVAL-5-160")
        pgp-pubalgorithm  =  "pubkey" "=" ( "rsa" | "rsa-encrypt"
                             | "rsa-sign" | "elgamal" | "dsa" | token )
        nonce             =  "nonce" "=" nonce-value
        nonce-value       =  quoted-string


   The meanings of the values of the parameters used above are as
   follows:

        realm: A string to be displayed to users so they know which
             identity to use. This string SHOULD contain at least the
             name of the host performing the authentication and MAY



Handley/Schulzrinne/Schooler/Rosenberg                      [Page 126]

Internet Draft                    SIP                      July 13, 2000


             additionally indicate the collection of users who might
             have access. An example might be "Users with call-out
             privileges".

        pgp-micalgorithm: The value of this parameter indicates the PGP
             message integrity check (MIC) to be used to produce the
             signature. If this parameter is not present, it is assumed
             to be "md5".  The currently defined values are "md5" for
             the MD5 checksum, and "sha1" for the SHA.1 algorithm.

        pgp-pubalgorithm: The value of this parameter indicates the PGP
             public-key algorithm to be used for signing and encrypting
             messages. If this parameter is not present, it is assumed
             to be "rsa" for the RSA algorithm. The value "dsa" defines
             the DSS/DH algorithm.

        pgp-version: The version of PGP that the client MUST use. Common
             values are "2.6.2" and "5.0". The default is 5.0.

        nonce: A server-specified data string which should be uniquely
             generated each time a 401 response is made. It is
             RECOMMENDED that this string be base64 [46] or hexadecimal
             data. Specifically, since the string is passed in the
             header lines as a quoted string, the double-quote character
             is not allowed. The contents of the nonce are
             implementation dependent. The quality of the implementation
             depends on a good choice. Since the nonce is used only to
             prevent replay attacks and is signed, a time stamp in units
             convenient to the server is sufficient.


             Replay attacks within the duration of the call setup
             are of limited interest, so that timestamps with a
             resolution of a few seconds are often sufficient. In
             that case, the server does not have to keep a record
             of the nonces.

   Example:

   WWW-Authenticate: pgp version="5.0"
     realm="Your Startrek identity, please", algorithm=md5,
     nonce="913082051"



15.1.2 The Authorization Request Header

   The client is expected to retry the request, passing an Authorization



Handley/Schulzrinne/Schooler/Rosenberg                      [Page 127]

Internet Draft                    SIP                      July 13, 2000


   header line, which is defined as follows.



        Authorization  =  "Authorization" ":" "pgp" # pgp-response
        pgp-response   =  realm | pgp-version | pgp-signature
                          | signed-by | nonce
        pgp-signature  =  "signature" "=" quoted-string
        signed-by      =  "signed-by" "=" <"> URI <">


   The client MUST increment the CSeq header before resubmitting the
   request. The signature MUST correspond to the From header of the
   request unless the signed-by parameter is provided.

        pgp-signature: The PGP ASCII-armored signature [42] and [45], as
             it appears between the "BEGIN PGP MESSAGE" and "END PGP
             MESSAGE" delimiters, without the version indication. The
             signature is included without any linebreaks.

             The signature is computed, in order, across the nonce (if
             present), realm, request method, request version and header
             fields following the Authorization header and the message
             body, in the same order as they appear in the message. The
             nonce, realm, request method and version are prepended to
             the header fields without any white space. The signature is
             computed across the headers as sent, and the terminating
             CRLF. The CRLF following the Authorization header is NOT
             included in the signature.

             UACs MAY attempt to authenticate themselves without a nonce
             on the first INVITE request rather than waiting for a 401
             response if the UAC knows, e.g., from past requests or
             local configuration, that the UAS supports PGP
             authentication. A server MAY be configured not to generate
             nonces only if replay attacks are not a concern.


             Not generating nonces avoids the additional set of
             request, 401 response and possibly ACK messages and
             reduces delay by one round-trip time.


             Using the ASCII-armored version is about 25% less
             space-efficient than including the binary signature,
             but it is significantly easier for the receiver to
             piece together. Versions of the PGP program always
             include the full (compressed) signed text in their



Handley/Schulzrinne/Schooler/Rosenberg                      [Page 128]

Internet Draft                    SIP                      July 13, 2000


             output unless ASCII-armored mode ( -sta ) is
             specified.  Typical signatures are about 200 bytes
             long. -- The PGP signature mechanism allows the client
             to simply pass the request to an external PGP program.
             This relies on the requirement that proxy servers are
             not allowed to reorder or change header fields.

        realm: The realm is copied from the corresponding WWW-
             Authenticate header field parameter.

        signed-by: If and only if the request was not signed by the
             entity listed in the From header, the signed-by header
             indicates the name of the signing entity, expressed as a
             URI.

   Receivers of signed SIP messages SHOULD discard any end-to-end header
   fields above the Authorization header, as they may have been
   maliciously added en route by a proxy.

   Example:

   Authorization: pgp version="5.0",
     realm="Your Startrek identity, please",
     nonce="913082051"
     signature="iQB1AwUBNNJiUaYBnHmiiQh1AQFYsgL/Wt3dk6TWK81/b0gcNDf
     VAUGU4rhEBW972IPxFSOZ94L1qhCLInTPaqhHFw1cb3lB01rA0RhpV4t5yCdUt
     SRYBSkOK29o5e1KlFeW23EzYPVUm2TlDAhbcjbMdfC+KLFX
     =aIrx"



15.2 PGP Encryption Scheme

   The PGP encryption scheme uses the following syntax:



        Encryption    =  "Encryption" ":" "pgp" pgp-eparams
        pgp-eparams   =  1# ( pgp-version | pgp-encoding )
        pgp-encoding  =  "encoding" "=" "ascii" | token


        encoding: Describes the encoding or "armor" used by PGP. The
             value "ascii" refers to the standard PGP ASCII armor,
             without the lines containing "BEGIN PGP MESSAGE" and "END
             PGP MESSAGE" and without the version identifier. By
             default, the encrypted part is included as binary.




Handley/Schulzrinne/Schooler/Rosenberg                      [Page 129]

Internet Draft                    SIP                      July 13, 2000


   Example:

   Encryption: pgp version="2.6.2", encoding="ascii"



15.3 Response-Key Header Field for PGP



        Response-Key  =  "Response-Key" ":" "pgp" pgp-eparams
        pgp-eparams   =  1# ( pgp-version | pgp-encoding | pgp-key)
        pgp-key       =  "key" "=" quoted-string


   If ASCII encoding has been requested via the encoding parameter, the
   key parameter contains the user's public key as extracted from the
   pgp key ring with the "pgp -kxa user ".

   Example:

   Response-Key: pgp version="2.6.2", encoding="ascii",
     key="mQBtAzNWHNYAAAEDAL7QvAdK2utY05wuUG+ItYK5tCF8HNJM60sU4rLaV+eUnkMk
     mOmJWtc2wXcZx1XaXb2lkydTQOesrUR75IwNXBuZXPEIMThEa5WLsT7VLme7njnx
     sE86SgWmAZx5ookIdQAFEbQxSGVubmluZyBTY2h1bHpyaW5uZSA8c2NodWx6cmlu
     bmVAY3MuY29sdW1iaWEuZWR1Pg==
     =+y19"



16 Examples

   In the following examples, we often omit the message body and the
   corresponding Content-Length and Content-Type headers for brevity.

16.1 Registration

   A user at host saturn.bell-tel.com registers on start-up, via
   multicast, with the local SIP server named bell-tel.com. In the
   example, the user agent on saturn expects to receive SIP requests on
   UDP port 3890.


   C->S: REGISTER sip:bell-tel.com SIP/2.0
         Via: SIP/2.0/UDP saturn.bell-tel.com
         From: sip:watson@bell-tel.com
         To: sip:watson@bell-tel.com
         Call-ID: 70710@saturn.bell-tel.com



Handley/Schulzrinne/Schooler/Rosenberg                      [Page 130]

Internet Draft                    SIP                      July 13, 2000


         CSeq: 1 REGISTER
         Contact: 
         Expires: 7200



   The registration expires after two hours. Any future invitations for
   watson@bell-tel.com arriving at sip.bell-tel.com will now be
   redirected to watson@saturn.bell-tel.com, UDP port 3890.

   If Watson wants to be reached elsewhere, say, an on-line service he
   uses while traveling, he updates his reservation after first
   cancelling any existing locations:


   C->S: REGISTER sip:bell-tel.com SIP/2.0
         Via: SIP/2.0/UDP saturn.bell-tel.com
         From: sip:watson@bell-tel.com
         To: sip:watson@bell-tel.com
         Call-ID: 70710@saturn.bell-tel.com
         CSeq: 2 REGISTER
         Contact: *
         Expires: 0

   C->S: REGISTER sip:bell-tel.com SIP/2.0
         Via: SIP/2.0/UDP saturn.bell-tel.com
         From: sip:watson@bell-tel.com
         To: sip:watson@bell-tel.com
         Call-ID: 70710@saturn.bell-tel.com
         CSeq: 3 REGISTER
         Contact: sip:tawatson@example.com



   Now, the server will forward any request for Watson to the server at
   example.com, using the Request-URI tawatson@example.com. For the
   server at example.com to reach Watson, he will need to send a
   REGISTER there, or inform the server of his current location through
   some other means.

   It is possible to use third-party registration. Here, the secretary
   jon.diligent registers his boss, T. Watson:

   C->S: REGISTER sip:bell-tel.com SIP/2.0
         Via: SIP/2.0/UDP pluto.bell-tel.com
         From: sip:jon.diligent@bell-tel.com
         To: sip:watson@bell-tel.com
         Call-ID: 17320@pluto.bell-tel.com



Handley/Schulzrinne/Schooler/Rosenberg                      [Page 131]

Internet Draft                    SIP                      July 13, 2000


         CSeq: 1 REGISTER
         Contact: sip:tawatson@example.com



   The request could be sent to either the registrar at bell-tel.com or
   the server at example.com. In the latter case, the server at
   example.com would proxy the request to the address indicated in the
   Request-URI. Then, Max-Forwards header could be used to restrict the
   registration to that server.

16.2 Invitation to a Multicast Conference

   The first example invites bob@example.com to a multicast session.
   All examples use the Session Description Protocol (SDP) (RFC 2327
   [6]) as the session description format.

16.2.1 Request


   C->S: INVITE sip:bob@one.example.com SIP/2.0
         Via: SIP/2.0/UDP sip.example.com;branch=7c337f30d7ce.1
           ;maddr=239.128.16.254;ttl=16
         Via: SIP/2.0/UDP mouse.wonderland.com
         From: Alice 
         To: Bob 
         Call-ID: 602214199@mouse.wonderland.com
         CSeq: 1 INVITE
         Contact: Alice 
         Subject: SIP will be discussed, too
         Content-Type: application/sdp
         Content-Length: 187

         v=0
         o=user1 53655765 2353687637 IN IP4 128.3.4.5
         s=Mbone Audio
         t=3149328700 0
         i=Discussion of Mbone Engineering Issues
         e=mbone@somewhere.com
         c=IN IP4 224.2.0.1/127
         t=0 0
         m=audio 3456 RTP/AVP 0
         a=rtpmap:0 PCMU/8000



   The From request header above states that the request was initiated
   by alice@wonderland.com and addressed to bob@example.com (From header



Handley/Schulzrinne/Schooler/Rosenberg                      [Page 132]

Internet Draft                    SIP                      July 13, 2000


   fields). The Via fields list the hosts along the path from invitation
   initiator (the last element of the list) towards the callee. In the
   example above, the message was last multicast to the administratively
   scoped group 239.128.16.254 with a ttl of 16 from the host
   sip.example.com. The second Via header field indicates that it was
   originally sent from the outbound proxy mouse.wonderland.com. The
   Request-URI indicates that the request is currently being being
   addressed to bob@one.example.com, the local address that the SIP
   server for the example.com domain looked up for the callee.

   In this case, the session description is using the Session
   Description Protocol (SDP), as stated in the Content-Type header.

   The header is terminated by an empty line and is followed by a
   message body containing the session description.

16.2.2 Response

   The called user agent, directly or indirectly through proxy servers,
   indicates that it is alerting ("ringing") the called party:


   S->C: SIP/2.0 180 Ringing
         Via: SIP/2.0/UDP csvax.cs.caltech.edu;branch=7c337f30d7ce.1
           ;maddr=239.128.16.254;ttl=16
         Via: SIP/2.0/UDP north.east.isi.edu
         From: Alice 
         To: Bob  ;tag=3141593
         Call-ID: 602214199@mouse.wonderland.com
         CSeq: 1 INVITE



   A sample response to the invitation is given below. The first line of
   the response states the SIP version number, that it is a 200 (OK)
   response, which means the request was successful. The Via headers are
   taken from the request, and entries are removed hop by hop as the
   response retraces the path of the request. A new authentication field
   MAY be added by the invited user's agent if required. The Call-ID is
   taken directly from the original request, along with the remaining
   fields of the request message. The original sense of From field is
   preserved (i.e., it is the session initiator).

   In addition, the Contact header gives details of the host where the
   user was located, or alternatively the relevant proxy contact point
   which should be reachable from the caller's host.





Handley/Schulzrinne/Schooler/Rosenberg                      [Page 133]

Internet Draft                    SIP                      July 13, 2000


   S->C: SIP/2.0 200 OK
         Via: SIP/2.0/UDP csvax.cs.caltech.edu;branch=7c337f30d7ce.1
           ;maddr=239.128.16.254;ttl=16
         Via: SIP/2.0/UDP north.east.isi.edu
         From: Alice 
         To: Bob  ;tag=3141593
         Call-ID: 602214199@mouse.wonderland.com
         CSeq: 1 INVITE
         Contact: 



   The caller confirms the invitation by sending an ACK request to the
   location named in the Contact header:


   C->S: ACK sip:bob@one.example.com SIP/2.0
         Via: SIP/2.0/UDP north.east.isi.edu
         From: Alice 
         To: Bob  ;tag=3141593
         Call-ID: 602214199@mouse.wonderland.com
         CSeq: 1 ACK



16.3 Two-party Call

   For two-party Internet phone calls, the response must contain a
   description of where to send the data. In the example below, Bell
   calls Watson. Bell indicates that he can receive RTP audio codings 0
   (PCMU), 3 (GSM), 4 (G.723) and 5 (DVI4).


   C->S: INVITE sip:watson@boston.bell-tel.com SIP/2.0
         Via: SIP/2.0/UDP kton.bell-tel.com
         From: A. Bell 
         To: T. Watson 
         Call-ID: 662606876@kton.bell-tel.com
         CSeq: 1 INVITE
         Contact: 
         Subject: Mr. Watson, come here.
         Content-Type: application/sdp
         Content-Length: ...

         v=0
         o=bell 53655765 2353687637 IN IP4 128.3.4.5
         s=Mr. Watson, come here.
         t=3149328600 0



Handley/Schulzrinne/Schooler/Rosenberg                      [Page 134]

Internet Draft                    SIP                      July 13, 2000


         c=IN IP4 kton.bell-tel.com
         m=audio 3456 RTP/AVP 0 3 4 5
         a=rtpmap:0 PCMU/8000
         a=rtpmap:3 GSM/8000
         a=rtpmap:4 G723/8000
         a=rtpmap:5 DVI4/8000

   S->C: SIP/2.0 100 Trying
         Via: SIP/2.0/UDP kton.bell-tel.com
         From: A. Bell 
         To: T. Watson  ;tag=37462311
         Call-ID: 662606876@kton.bell-tel.com
         CSeq: 1 INVITE
         Content-Length: 0

   S->C: SIP/2.0 180 Ringing
         Via: SIP/2.0/UDP kton.bell-tel.com
         From: A. Bell 
         To: T. Watson  ;tag=37462311
         Call-ID: 662606876@kton.bell-tel.com
         CSeq: 1 INVITE
         Content-Length: 0

   S->C: SIP/2.0 182 Queued, 2 callers ahead
         Via: SIP/2.0/UDP kton.bell-tel.com
         From: A. Bell 
         To: T. Watson  ;tag=37462311
         Call-ID: 3298420296@kton.bell-tel.com
         CSeq: 1 INVITE
         Content-Length: 0

   S->C: SIP/2.0 182 Queued, 1 caller ahead
         Via: SIP/2.0/UDP kton.bell-tel.com
         From: A. Bell 
         To: T. Watson  ;tag=37462311
         Call-ID: 3298420296@kton.bell-tel.com
         CSeq: 1 INVITE
         Content-Length: 0

   S->C: SIP/2.0 200 OK
         Via: SIP/2.0/UDP kton.bell-tel.com
         From: A. Bell 
         To:  ;tag=37462311
         Call-ID: 3298420296@kton.bell-tel.com
         CSeq: 1 INVITE
         Contact: sip:watson@boston.bell-tel.com
         Content-Type: application/sdp
         Content-Length: ...



Handley/Schulzrinne/Schooler/Rosenberg                      [Page 135]

Internet Draft                    SIP                      July 13, 2000


         v=0
         o=watson 4858949 4858949 IN IP4 192.1.2.3
         s=I'm on my way
         t=3149329600 0
         c=IN IP4 boston.bell-tel.com
         m=audio 5004 RTP/AVP 0 3
         a=rtpmap:0 PCMU/8000
         a=rtpmap:3 GSM/8000



   The example illustrates the use of informational status responses.
   Here, the reception of the call is confirmed immediately (100), then,
   possibly after some database mapping delay, the call rings (180) and
   is then queued, with periodic status updates.

   Watson can only receive PCMU and GSM. Note that Watson's list of
   codecs may or may not be a subset of the one offered by Bell, as each
   party indicates the data types it is willing to receive. Watson will
   send audio data to port 3456 at c.bell-tel.com, Bell will send to
   port 5004 at boston.bell-tel.com.

   By default, the media session is one RTP session. Watson will receive
   RTCP packets on port 5005, while Bell will receive them on port 3457.

   Since the two sides have agreed on the set of media, Bell confirms
   the call without enclosing another session description:


   C->S: ACK sip:watson@boston.bell-tel.com SIP/2.0
         Via: SIP/2.0/UDP kton.bell-tel.com
         From: A. Bell 
         To: T. Watson  ;tag=37462311
         Call-ID: 3298420296@kton.bell-tel.com
         CSeq: 1 ACK



16.4 Terminating a Call

   To terminate a call, caller or callee can send a BYE request:


   C->S: BYE sip:watson@boston.bell-tel.com SIP/2.0
         Via: SIP/2.0/UDP kton.bell-tel.com
         From: A. Bell 
         To: T. A. Watson  ;tag=37462311
         Call-ID: 3298420296@kton.bell-tel.com



Handley/Schulzrinne/Schooler/Rosenberg                      [Page 136]

Internet Draft                    SIP                      July 13, 2000


         CSeq: 2 BYE



   If the callee wants to abort the call, it simply reverses the To and
   From fields. Note that it is unlikely that a BYE from the callee will
   traverse the same proxies as the original INVITE.

16.5 Forking Proxy

   In this example, Bell (a.g.bell@bell-tel.com) (C), currently seated
   at host c.bell-tel.com wants to call Watson (t.watson@ieee.org). At
   the time of the call, Watson is logged in at two workstations,
   t.watson@x.bell-tel.com (X) and watson@y.bell-tel.com (Y), and has
   registered with the IEEE proxy server (P) called sip.ieee.org. The
   IEEE server also has a registration for the home machine of Watson,
   at watson@h.bell-tel.com (H), as well as a permanent registration at
   watson@acm.org (A). For brevity, the examples omit the message bodies
   containing the session descriptions.

   Bell's user agent sends the invitation to the SIP server for the
   ieee.org domain:


   C->P: INVITE sip:t.watson@ieee.org SIP/2.0
         Via:     SIP/2.0/UDP c.bell-tel.com
         From:    A. Bell 
         To:      T. Watson 
         Call-ID: 31415@c.bell-tel.com
         CSeq:    1 INVITE
         Contact: a.g.bell@c.bell-tel.com



   The SIP server at ieee.org tries the four addresses in parallel.  It
   sends the following message to the home machine:


   P->H: INVITE sip:watson@h.bell-tel.com SIP/2.0
         Via:     SIP/2.0/UDP sip.ieee.org ;branch=3d8a50dbf5a28d.1
         Via:     SIP/2.0/UDP c.bell-tel.com
         From:    A. Bell 
         To:      T. Watson 
         Call-ID: 31415@c.bell-tel.com
         CSeq:    1 INVITE
         Contact: a.g.bell@c.bell-tel.com





Handley/Schulzrinne/Schooler/Rosenberg                      [Page 137]

Internet Draft                    SIP                      July 13, 2000


   This request immediately yields a 404 (Not Found) response, since
   Watson is not currently logged in at home:


   H->P: SIP/2.0 404 Not Found
         Via:     SIP/2.0/UDP sip.ieee.org ;branch=3d8a50dbf5a28d.1
         Via:     SIP/2.0/UDP c.bell-tel.com
         From:    A. Bell 
         To:      T. Watson ;tag=87454273
         Call-ID: 31415@c.bell-tel.com
         CSeq:    1 INVITE



   The proxy ACKs the response so that host H can stop retransmitting
   it:

   P->H: ACK sip:watson@h.bell-tel.com SIP/2.0
         Via:     SIP/2.0/UDP sip.ieee.org ;branch=3d8a50dbf5a28d.1
         From:    A. Bell 
         To:      T. Watson ;tag=87454273
         Call-ID: 31415@c.bell-tel.com
         CSeq:    1 ACK



   Also, P attempts to reach Watson through the ACM server:

   P->A: INVITE sip:watson@acm.org SIP/2.0
         Via:     SIP/2.0/UDP sip.ieee.org ;branch=3d8a50dbf5a28d.2
         Via:     SIP/2.0/UDP c.bell-tel.com
         From:    A. Bell 
         To:      T. Watson 
         Call-ID: 31415@c.bell-tel.com
         CSeq:    1 INVITE
         Contact: a.g.bell@c.bell-tel.com



   In parallel, the next attempt proceeds, with an INVITE to X and Y:


   P->X: INVITE sip:t.watson@x.bell-tel.com SIP/2.0
         Via:     SIP/2.0/UDP sip.ieee.org ;branch=3d8a50dbf5a28d.3
         Via:     SIP/2.0/UDP c.bell-tel.com
         From:    A. Bell 
         To:      T. Watson 
         Call-ID: 31415@c.bell-tel.com



Handley/Schulzrinne/Schooler/Rosenberg                      [Page 138]

Internet Draft                    SIP                      July 13, 2000


         CSeq:    1 INVITE
         Contact: a.g.bell@c.bell-tel.com

   P->Y: INVITE sip:watson@y.bell-tel.com SIP/2.0
         Via:     SIP/2.0/UDP sip.ieee.org ;branch=3d8a50dbf5a28d.4
         Via:     SIP/2.0/UDP c.bell-tel.com
         From:    A. Bell 
         To:      T. Watson 
         Call-ID: 31415@c.bell-tel.com
         CSeq:    1 INVITE
         Contact: a.g.bell@c.bell-tel.com



   As it happens, both Watson at X and a colleague in the other lab at
   host Y hear the phones ringing and pick up. Both X and Y return 200s
   via the proxy to Bell.


   X->P: SIP/2.0 200 OK
         Via:      SIP/2.0/UDP sip.ieee.org ;branch=3d8a50dbf5a28d.3
         Via:      SIP/2.0/UDP c.bell-tel.com
         From:     A. Bell 
         To:       T. Watson  ;tag=192137601
         Call-ID:  31415@c.bell-tel.com
         CSeq:     1 INVITE
         Contact:  sip:t.watson@x.bell-tel.com

   Y->P: SIP/2.0 200 OK
         Via:      SIP/2.0/UDP sip.ieee.org ;branch=3d8a50dbf5a28d.4
         Via:      SIP/2.0/UDP c.bell-tel.com
         Contact:  sip:t.watson@y.bell-tel.com
         From:     A. Bell 
         To:       T. Watson  ;tag=35253448
         Call-ID:  31415@c.bell-tel.com
         CSeq:     1 INVITE



   Both responses are forwarded to Bell, using the Via information.  At
   this point, the ACM server is still searching its database. P can now
   cancel this attempt:


   P->A: CANCEL sip:watson@acm.org SIP/2.0
         Via:     SIP/2.0/UDP sip.ieee.org ;branch=3d8a50dbf5a28d.2
         From:    A. Bell 
         To:      T. Watson 



Handley/Schulzrinne/Schooler/Rosenberg                      [Page 139]

Internet Draft                    SIP                      July 13, 2000


         Call-ID: 31415@c.bell-tel.com
         CSeq:    1 CANCEL



   The ACM server gladly stops its neural-network database search and
   responds with a 200. The 200 will not travel any further, since P is
   the last Via stop.


   A->P: SIP/2.0 200 OK
         Via:     SIP/2.0/UDP sip.ieee.org ;branch=3d8a50dbf5a28d.2
         From:    A. Bell 
         To:      T. Watson 
         Call-ID: 31415@c.bell-tel.com
         CSeq:    1 CANCEL



   In addition, P responds to the original INVITE request with a 487
   (Request Terminated):

   A->P: SIP/2.0 487 Request Terminated
         Via:     SIP/2.0/UDP sip.ieee.org ;branch=3d8a50dbf5a28d.2
         Via:     SIP/2.0/UDP c.bell-tel.com
         From:    A. Bell 
         To:      T. Watson 
         Call-ID: 31415@c.bell-tel.com
         CSeq:    1 INVITE


   This response terminates at P.

   Bell gets the two 200 responses from X and Y in short order. Bell's
   reaction now depends on his software. He can either send an ACK to
   both if human intelligence is needed to determine who he wants to
   talk to or he can automatically reject one of the two calls. Here, he
   acknowledges both, separately and directly to the final destination:


   C->X: ACK sip:t.watson@x.bell-tel.com SIP/2.0
         Via:      SIP/2.0/UDP c.bell-tel.com
         From:     A. Bell 
         To:       T. Watson ;tag=192137601
         Call-ID:  31415@c.bell-tel.com
         CSeq:     1 ACK

   C->Y: ACK sip:watson@y.bell-tel.com SIP/2.0



Handley/Schulzrinne/Schooler/Rosenberg                      [Page 140]

Internet Draft                    SIP                      July 13, 2000


         Via:      SIP/2.0/UDP c.bell-tel.com
         From:     A. Bell 
         To:       T. Watson ;tag=35253448
         Call-ID:  31415@c.bell-tel.com
         CSeq:     1 ACK



   After a brief discussion between Bell with X and Y, it becomes clear
   that Watson is at X. (Note that this is not a three-way call; only
   Bell can talk to X and Y, but X and Y cannot talk to each other.)
   Thus, Bell sends a BYE to Y, which is replied to:


   C->Y: BYE sip:watson@y.bell-tel.com SIP/2.0
         Via:      SIP/2.0/UDP c.bell-tel.com
         From:     A. Bell 
         To:       T. Watson ;tag=35253448
         Call-ID:  31415@c.bell-tel.com
         CSeq:     2 BYE

   Y->C: SIP/2.0 200 OK
         Via:      SIP/2.0/UDP c.bell-tel.com
         From:     A. Bell 
         To:       T. Watson ;tag=35253448
         Call-ID:  31415@c.bell-tel.com
         CSeq:     2 BYE



16.6 Redirects

   Replies with status codes 301 (Moved Permanently) or 302 (Moved
   Temporarily) specify another location using the Contact field.
   Continuing our earlier example, the server P at ieee.org decides to
   redirect rather than proxy the request:


   P->C: SIP/2.0 302 Moved temporarily
         Via:     SIP/2.0/UDP c.bell-tel.com
         From:    A. Bell 
         To:      T. Watson ;tag=72538263
         Call-ID: 31415@c.bell-tel.com
         CSeq:    1 INVITE
         Contact: sip:watson@h.bell-tel.com,
                   sip:watson@acm.org, sip:t.watson@x.bell-tel.com,
                   sip:watson@y.bell-tel.com




Handley/Schulzrinne/Schooler/Rosenberg                      [Page 141]

Internet Draft                    SIP                      July 13, 2000


   As another example, assume Alice (A) wants to delegate her calls to
   Bob (B) while she is on vacation until July 29th, 1998. Any calls
   meant for her will reach Bob with Alice's To field, indicating to him
   what role he is to play. Charlie (C) calls Alice (A), whose server
   returns:


   A->C: SIP/2.0 302 Moved temporarily
         From: Charlie 
         To: Alice  ;tag=2332462
         Call-ID: 27182@caller.com
         Contact: sip:bob@anywhere.com
         Expires: Wed, 29 Jul 1998 9:00:00 GMT
         CSeq: 1 INVITE



   Charlie then sends the following request to the SIP server of the
   anywhere.com domain. Note that the server at anywhere.com forwards
   the request to Bob based on the Request-URI.


   C->B: INVITE sip:bob@anywhere.com SIP/2.0
         From: sip:charlie@caller.com
         To: sip:alice@anywhere.com
         Call-ID: 27182@caller.com
         CSeq: 2 INVITE
         Contact: sip:charlie@h.caller.com



   In the third redirection example, we assume that all outgoing
   requests are directed through a local firewall F at caller.com, with
   Charlie again inviting Alice:


   C->F: INVITE sip:alice@anywhere.com SIP/2.0
         From: sip:charlie@caller.com
         To: Alice 
         Call-ID: 27182@caller.com
         CSeq: 1 INVITE
         Contact: sip:charlie@h.caller.com



   The local firewall at caller.com happens to be overloaded and thus
   redirects the call from Charlie to a secondary server S:




Handley/Schulzrinne/Schooler/Rosenberg                      [Page 142]

Internet Draft                    SIP                      July 13, 2000


   F->C: SIP/2.0 302 Moved temporarily
         From: sip:charlie@caller.com
         To: Alice 
         Call-ID: 27182@caller.com
         CSeq: 1 INVITE
         Contact: 



   Based on this response, Charlie directs the same invitation to the
   secondary server spare.caller.com at port 5080, but maintains the
   same Request-URI as before:


   C->S: INVITE sip:alice@anywhere.com SIP/2.0
         From: sip:charlie@caller.com
         To: Alice 
         Call-ID: 27182@caller.com
         CSeq: 2 INVITE
         Contact: sip:charlie@h.caller.com



16.7 Negotiation

   An example of a 606 (Not Acceptable) response is:


   S->C: SIP/2.0 606 Not Acceptable
         From: sip:mjh@isi.edu
         To:  ;tag=7434264
         Call-ID: 14142@north.east.isi.edu
         CSeq: 1 INVITE
         Contact: sip:mjh@north.east.isi.edu
         Warning: 370 "Insufficient bandwidth (only have ISDN)",
           305 "Incompatible media format",
           330 "Multicast not available"
         Content-Type: application/sdp
         Content-Length: 50

         v=0
         o=schooler 3149329138 3149329165 IN IP4 38.245.76.2
         s=Let's talk
         t=3149328630 0
         b=CT:128
         c=IN IP4 north.east.isi.edu
         m=audio 3456 RTP/AVP 5 0 7
         a=rtpmap:5 DVI4/8000



Handley/Schulzrinne/Schooler/Rosenberg                      [Page 143]

Internet Draft                    SIP                      July 13, 2000


         a=rtpmap:0 PCMU/8000
         a=rtpmap:7 LPC/8000
         m=video 2232 RTP/AVP 31
         a=rtpmap:31 H261/90000



   In this example, the original request specified a bandwidth that was
   higher than the access link could support, requested multicast, and
   requested a set of media encodings. The response states that only 128
   kb/s is available and that (only) DVI, PCM or LPC audio could be
   supported in order of preference.

   The response also states that multicast is not available.  In such a
   case, it might be appropriate to set up a transcoding gateway and
   re-invite the user.

16.8 OPTIONS Request

   A caller Alice can use an OPTIONS request to find out the
   capabilities of a potential callee Bob, without "ringing" the
   designated address. Bob returns a description indicating that he is
   capable of receiving audio encodings PCM mu-law (RTP payload type 0),
   1016 (payload type 1), GSM (payload type 3), and SX7300/8000 (dynamic
   payload type 99), and video encodings H.261 (payload type 31) and
   H.263 (payload type 34).


   C->S: OPTIONS sip:bob@example.com SIP/2.0
         From: Alice 
         To: Bob 
         Call-ID: 6378@host.anywhere.org
         CSeq: 1 OPTIONS
         Accept: application/sdp

   S->C: SIP/2.0 200 OK
         From: Alice 
         To: Bob  ;tag=376364382
         Call-ID: 6378@host.anywhere.org
         Content-Length: 81
         Content-Type: application/sdp

         v=0
         o=alice 3149329138 3149329165 IN IP4 24.124.37.3
         s=Security problems
         t=3149328650 0
         c=IN IP4 24.124.37.3
         m=audio 0 RTP/AVP 0 1 3 99



Handley/Schulzrinne/Schooler/Rosenberg                      [Page 144]

Internet Draft                    SIP                      July 13, 2000


         a=rtpmap:0 PCMU/8000
         a=rtpmap:1 1016/8000
         a=rtpmap:3 GSM/8000
         a=rtpmap:99 SX7300/8000
         m=video 0 RTP/AVP 31 34
         a=rtpmap:31 H261/90000
         a=rtpmap:34 H263/90000



A Minimal Implementation

A.1 Transport Protocol Support

   User agents and stateless proxies MUST support UDP and MAY support
   TCP or other transport protocols, stateful proxies MUST support both
   UDP and TCP.

A.2 Client

   All clients MUST be able to generate the INVITE and ACK requests.
   Clients MUST generate and parse the Call-ID, Content-Length,
   Content-Type, CSeq, From and To headers. Clients MUST also parse the
   Require header. A minimal implementation MUST understand SDP (RFC
   2327, [6]). It MUST be able to recognize the status code classes 1
   through 6 and act accordingly.

   The following capability sets build on top of the minimal
   implementation described in the previous paragraph. In general, each
   capability listed below builds on the ones above it:

        Basic: A basic implementation adds support for the BYE method to
             allow the interruption of a pending call attempt. It
             includes a User-Agent header in its requests and indicates
             its preferred language in the Accept-Language header.

        Redirection: To support call forwarding, a client needs to be
             able to understand the Contact header, but only the SIP-URL
             part, not the parameters.

        Firewall-friendly: A firewall-friendly client understands the
             Route and Record-Route header fields and can be configured
             to use a local proxy for all outgoing requests.

        Negotiation: A client MUST be able to request the OPTIONS method
             and understand the 380 (Alternative Service) status and the
             Contact parameters to participate in terminal and media
             negotiation. It SHOULD be able to parse the Warning



Handley/Schulzrinne/Schooler/Rosenberg                      [Page 145]

Internet Draft                    SIP                      July 13, 2000


             response header to provide useful feedback to the caller.

        Authentication: If a client wishes to invite callees that
             require caller authentication, it MUST be able to recognize
             the 401 (Unauthorized) status code, MUST be able to
             generate the Authorization request header and MUST
             understand the WWW-Authenticate response header.

             If a client wishes to use proxies that require caller
             authentication, it MUST be able to recognize the 407 (Proxy
             Authentication Required) status code, MUST be able to
             generate the Proxy-Authorization request header and
             understand the Proxy-Authenticate response header.

A.3 Server

   A minimally compliant server implementation MUST understand the
   INVITE, ACK, OPTIONS and BYE requests. A proxy server MUST also
   understand CANCEL. It MUST parse and generate, as appropriate, the
   Call-ID, Content-Length, Content-Type, CSeq, Expires, From, Max-
   Forwards, Require, To and Via headers. It MUST echo the CSeq and
   Timestamp headers in the response. It SHOULD include the Server
   header in its responses.

A.4 Header Processing

   Table 6 lists the headers that different implementations support. UAC
   refers to a user-agent client (calling user agent), UAS to a user-
   agent server (called user-agent).

   The fields in the table have the following meaning. Type is as in
   Table 4 and 5. "-" indicates the field is not meaningful to this
   system (although it might be generated by it). "m" indicates the
   field MUST be understood. "b" indicates the field SHOULD be
   understood by a Basic implementation.  "r" indicates the field SHOULD
   be understood if the system claims to understand redirection. "a"
   indicates the field SHOULD be understood if the system claims to
   support authentication. "e" indicates the field SHOULD be understood
   if the system claims to support encryption. "o" indicates support of
   the field is purely optional. Headers whose support is optional for
   all implementations are not shown.


B Usage of the Session Description Protocol (SDP)

   This section describes the use of the Session Description Protocol
   (SDP) (RFC 2327 [6]). SDP is identified as Content-Type
   "application/sdp".  Each SIP message body may only contain one SDP



Handley/Schulzrinne/Schooler/Rosenberg                      [Page 146]

Internet Draft                    SIP                      July 13, 2000




                        type  UAC  proxy  UAS  registrar
   _____________________________________________________
   Accept                R     -     o     m      m
   Accept-Encoding       R     -     -     m      m
   Accept-Language       R     -     b     b      b
   Allow                405    o     -     -      -
   Authorization         R     a     o     a      a
   Call-ID               g     m     m     m      m
   Content-Encoding      g     m     -     m      m
   Content-Length        g     m     m     m      m
   Content-Type          g     m     -     m      m
   CSeq                  g     m     m     m      m
   Encryption            g     e     -     e      e
   Expires               g     -     o     o      m
   From                  g     m     o     m      m
   Hide                  R     -     m     -      -
   Contact               R     -     -     -      m
   Contact               r     r     r     -      -
   Max-Forwards          R     -     b     -      -
   Proxy-Authenticate   407    a     -     -      -
   Proxy-Authorization   R     -     a     -      -
   Proxy-Require         R     -     m     -      -
   Require               R     m     -     m      m
   Response-Key          R     -     -     e      e
   Route                 R     -     m     -      -
   Timestamp             g     o     o     m      m
   To                    g     m     m     m      m
   Unsupported           r     b     b     -      -
   User-Agent            g     b     -     b      -
   Via                   g     m     m     m      m
   WWW-Authenticate     401    a     -     -      -


   Table 6: Header Field Processing Requirements

   message unless the SIP message body is of type "multipart". SDP
   messages in later SIP request supersede earlier ones for the same
   call leg.

B.1 Configuring Media Streams

   The caller and callee align their media descriptions so that the nth
   media stream ("m=" line) in the caller's session description
   corresponds to the nth media stream in the callee's description.

   All media descriptions SHOULD contain "a=rtpmap" mappings from RTP
   payload types to encodings.


Handley/Schulzrinne/Schooler/Rosenberg                      [Page 147]

Internet Draft                    SIP                      July 13, 2000


        This allows easier migration away from static payload
        types.

   User agents MUST NOT reuse dynamic payload types for different
   encodings when issuing a re-INVITE.

   If the callee wants to neither send nor receive a stream offered by
   the caller, the callee sets the port number of that stream to zero in
   its media description.


        There currently is no other way than port zero for the
        callee to refuse a bidirectional stream offered by the
        caller. Both caller and callee need to be aware what media
        tools are to be started.

   For example, assume that the caller Alice has included the following
   description in her INVITE request. It includes a bidirectional audio
   stream and two bidirectional video streams, using H.261 (payload type
   31) and MPEG (payload type 32).


   v=0
   o=alice 2890844526 2890844526 IN IP4 host.anywhere.com
   s=New board design
   t=0 0
   c=IN IP4 host.anywhere.com
   m=audio 49170 RTP/AVP 0
   a=rtpmap:0 PCMU/8000
   m=video 51372 RTP/AVP 31
   a=rtpmap:31 H261/90000
   m=video 53000 RTP/AVP 32
   a=rtpmap:32 MPV/90000



   The callee, Bob, does not want to receive or send the first video
   stream, so it returns the media description below:

   v=0
   o=bob 2890844730 2890844730 IN IP4 host.example.com
   s=New board design
   t=0 0
   c=IN IP4 host.example.com
   m=audio 47920 RTP/AVP 0 1
   a=rtpmap:0 PCMU/8000
   m=video 0 RTP/AVP 31
   m=video 53000 RTP/AVP 32



Handley/Schulzrinne/Schooler/Rosenberg                      [Page 148]

Internet Draft                    SIP                      July 13, 2000


   a=rtpmap:32 MPV/90000



B.2 Setting SDP Values for Unicast

   If a session description from a caller contains a media stream which
   is listed as send (receive) only, it means that the caller is only
   willing to send (receive) this stream, not receive (send). The same
   is true for the callee.

   For receive-only and send-or-receive streams, the port number and
   address in the session description indicate where the media stream
   should be sent to by the recipient of the session description, either
   caller or callee.  For send-only RTP streams, the address and port
   number indicate where RTCP reports are to be sent. (RTCP reports are
   sent to the port number one higher than the number indicated.)

   The list of payload types for each media stream conveys two pieces of
   information, namely the set of codecs that the caller or callee is
   capable of sending or receiving, and the RTP payload type numbers
   used to identify those codecs. For receive-only or send-and-receive
   media streams, a caller SHOULD list all of the codecs it is capable
   of supporting in the session description in an INVITE or ACK. For
   send-only streams, the caller SHOULD indicate only those it wishes to
   send for this session. For receive-only streams, the payload type
   numbers indicate the value of the payload type field in RTP packets
   the caller is expecting to receive for that codec type. For send-only
   streams, the payload type numbers indicate the value of the payload
   type field in RTP packets the caller is planning to send for that
   codec type.  For send-and-receive streams, the payload type numbers
   indicate the value of the payload type field the caller expects to
   both send and receive.

   If a media stream is listed as receive-only by the caller, the callee
   lists, in the response, those codecs it intends to use from among the
   ones listed in the request. If a media stream is listed as send-only
   by the caller, the callee lists, in the response, those codecs it is
   willing to receive among the ones listed in the the request. If the
   media stream is listed as both send and receive, the callee lists
   those codecs it is capable of sending or receiving among the ones
   listed by the caller in the INVITE. The actual payload type numbers
   in the callee's session description corresponding to a particular
   codec MUST be the same as the caller's session description.

   If caller and callee have no media formats in common for a particular
   stream, the callee MUST return a session description containing the
   particular "m=" line, but with the port number set to zero, and no



Handley/Schulzrinne/Schooler/Rosenberg                      [Page 149]

Internet Draft                    SIP                      July 13, 2000


   payload types listed.

   If there are no media formats in common for all streams, the callee
   SHOULD return a 488 or 606 response, with a 304 Warning header field.

B.3 Multicast Operation

   The interpretation of send-only and receive-only for multicast media
   sessions differs from that for unicast sessions. For multicast,
   send-only means that the recipient of the session description (caller
   or callee) SHOULD only send media streams to the address and port
   indicated. Receive-only means that the recipient of the session
   description SHOULD only receive media on the address and port
   indicated.

   For multicast, receive and send multicast addresses are the same and
   all parties use the same port numbers to receive media data. If the
   session description provided by the caller is acceptable to the
   callee, the callee can choose not to include a session description or
   MAY echo the description in the response.

   A callee MAY, in the response, return a session description with some
   of the payload types removed, or port numbers set to zero (but no
   other value). This indicates to the caller that the callee does not
   support the given stream or media types which were removed. A callee
   MUST NOT change whether a given stream is send-only, receive-only, or
   send-and-receive.

   If a callee does not support multicast at all, it SHOULD return a 400
   status response and include a 330 Warning.

B.4 Delayed Media Streams

   In some cases, a caller may not know the set of media formats which
   it can support at the time it would like to issue an invitation. This
   is the case when the caller is actually a gateway to another protocol
   which performs media format negotiation after call setup. When this
   occurs, a caller MAY issue an INVITE with a session description that
   contains no media lines. The callee SHOULD interpret this to mean
   that the caller wishes to participate in a multimedia session
   described by the session description, but that the media streams are
   not yet known. The callee SHOULD return a session description
   indicating the streams and media formats it is willing to support,
   however. The caller MAY update the session description either in the
   ACK request or in a re-INVITE at a later time, once the streams are
   known.

B.5 Adding and Deleting Media Streams



Handley/Schulzrinne/Schooler/Rosenberg                      [Page 150]

Internet Draft                    SIP                      July 13, 2000


   To add a stream to an existing call leg, either party appends an
   additional "m" line to the previous session description when sending
   a re-INVITE.

   To remove a stream from a call leg, either party sets its port to
   zero in the session description when sending a re-INVITE.

   UAs receiving SDP SHOULD accept descriptions that are not aligned, in
   terms of "m=" lines, with earlier descriptions. If a UA receives such
   a description, it SHOULD line up "m=" lines by media type ("audio",
   "video", ...). If it receives a new description in an INVITE request,
   it MAY line up the "m=" lines by media type, address and port and
   position in the description and start or delete media streams
   accordingly.


        Receiving a new description in an INVITE request that omits
        media with zero ports or is otherwise different may occur
        if the UA crashed and restarted.

B.6 Putting Media Streams on Hold

   If a party in a call wants to put the other party "on hold", i.e.,
   request that it temporarily stops sending one or more media streams,
   a party re-invites the other by sending an INVITE request with a
   modified session description. The session description is the same as
   in the original invitation (or response), but the "c" destination
   addresses for the media streams to be put on hold are set to zero
   (0.0.0.0).

B.7 Subject and SDP "s=" Line

   The SDP "s=" line and the SIP Subject header field have different
   meanings when inviting to a multicast session. The session
   description line describes the subject of the multicast session,
   while the SIP Subject header field describes the reason for the
   invitation. The example in Section 16.2 illustrates this point. For
   invitations to two-party sessions, the SDP "s=" line MAY consist of a
   single space character (0x20).


        Unfortunately, SDP does not allow to leave the "s=" line
        empty.

B.8 The SDP "o=" Line

   The "o=" line MUST be present for all sessions, including two-party
   sessions, to allow re-use of SDP-based tools.



Handley/Schulzrinne/Schooler/Rosenberg                      [Page 151]

Internet Draft                    SIP                      July 13, 2000


C Summary of Augmented BNF

   All of the mechanisms specified in this document are described in
   both prose and an augmented Backus-Naur Form (BNF) similar to that
   used by RFC 822 [28] and RFC 2234 [47]. Implementors will need to be
   familiar with the notation in order to understand this specification.
   The augmented BNF includes the following constructs:



        name  =  definition


   The name of a rule is simply the name itself (without any enclosing
   "<" and ">") and is separated from its definition by the equal "="
   character. White space is only significant in that indentation of
   continuation lines is used to indicate a rule definition that spans
   more than one line. Certain basic rules are in uppercase, such as SP,
   LWS, HT, CRLF, DIGIT, ALPHA, etc. Angle brackets are used within
   definitions whenever their presence will facilitate discerning the
   use of rule names.


   "literal"


   Quotation marks surround literal text. Unless stated otherwise, the
   text is case-insensitive.


   rule1 | rule2


   Elements separated by a bar ("|") are alternatives, e.g., "yes | no"
   will accept yes or no.


   (rule1 rule2)


   Elements enclosed in parentheses are treated as a single element.
   Thus, "(elem (foo | bar) elem)" allows the token sequences "elem foo
   elem" and "elem bar elem".


   *rule





Handley/Schulzrinne/Schooler/Rosenberg                      [Page 152]

Internet Draft                    SIP                      July 13, 2000


   The character "*" preceding an element indicates repetition. The full
   form is "*element" indicating at least  and at most 
   occurrences of element. Default values are 0 and infinity so that
   "*(element)" allows any number, including zero; "1*element" requires
   at least one; and "1*2element" allows one or two.


   [rule]


   Square brackets enclose optional elements; "[foo bar]" is equivalent
   to "*1(foo bar)".


   N rule


   Specific repetition: "(element)" is equivalent to
   "*(element)"; that is, exactly  occurrences of (element).
   Thus 2DIGIT is a 2-digit number, and 3ALPHA is a string of three
   alphabetic characters.


   #rule


   A construct "#" is defined, similar to "*", for defining lists of
   elements. The full form is "# element" indicating at least 
   and at most  elements, each separated by one or more commas (",")
   and OPTIONAL linear white space (LWS). This makes the usual form of
   lists very easy; a rule such as



           ( *LWS element *( *LWS "," *LWS element ))


   can be shown as 1# element. Wherever this construct is used, null
   elements are allowed, but do not contribute to the count of elements
   present. That is, "(element), , (element)" is permitted, but counts
   as only two elements. Therefore, where at least one element is
   required, at least one non-null element MUST be present. Default
   values are 0 and infinity so that "#element" allows any number,
   including zero; "1#element" requires at least one; and "1#2element"
   allows one or two.


   ; comment



Handley/Schulzrinne/Schooler/Rosenberg                      [Page 153]

Internet Draft                    SIP                      July 13, 2000


   A semi-colon, set off some distance to the right of rule text, starts
   a comment that continues to the end of line. This is a simple way of
   including useful notes in parallel with the specifications.


   implied *LWS


   The grammar described by this specification is word-based. Except
   where noted otherwise, linear white space (LWS) can be included
   between any two adjacent words (token or quoted-string), and between
   adjacent tokens and separators, without changing the interpretation
   of a field. At least one delimiter (LWS and/or separators) MUST exist
   between any two tokens (for the definition of "token" below), since
   they would otherwise be interpreted as a single token.

C.1 Basic Rules

   The following rules are used throughout this specification to
   describe basic parsing constructs. The US-ASCII coded character set
   is defined by ANSI X3.4-1986.


        OCTET     =  %x00-ff ; any 8-bit sequence of data
        CHAR      =  %x00-7f ; any US-ASCII character (octets 0 - 127)
        upalpha   =  "A" | "B" | "C" | "D" | "E" | "F" | "G" | "H" | "I" |
                     "J" | "K" | "L" | "M" | "N" | "O" | "P" | "Q" | "R" |
                     "S" | "T" | "U" | "V" | "W" | "X" | "Y" | "Z"
        lowalpha  =  "a" | "b" | "c" | "d" | "e" | "f" | "g" | "h" | "i" |
                     "j" | "k" | "l" | "m" | "n" | "o" | "p" | "q" | "r" |
                     "s" | "t" | "u" | "v" | "w" | "x" | "y" | "z"
        alpha     =  lowalpha | upalpha
        DIGIT     =  "0" | "1" | "2" | "3" | "4" | "5" | "6" | "7" |
                     "8" | "9"
        alphanum  =  alpha | DIGIT
        CTL       =  %x00-1f | 0x7f ; (octets 0 -- 31) and DEL (127)
        CR        =  %d13 ; US-ASCII CR, carriage return character
        LF        =  %d10 ; US-ASCII LF, line feed character
        SP        =  %d32 ; US-ASCII SP, space character
        HT        =  %d09 ; US-ASCII HT, horizontal tab character
        CRLF      =  CR LF ; typically the end of a line


   The following are defined in RFC 2396 [12] for the SIP URI:


        unreserved  =  alphanum | mark
        mark        =  "-" | "_" | "." | "!" | "~" | "*" | "'"



Handley/Schulzrinne/Schooler/Rosenberg                      [Page 154]

Internet Draft                    SIP                      July 13, 2000


                   |   "(" | ")"
        escaped     =  "%" hex hex


   SIP header field values can be folded onto multiple lines if the
   continuation line begins with a space or horizontal tab. All linear
   white space, including folding, has the same semantics as SP. A
   recipient MAY replace any linear white space with a single SP before
   interpreting the field value or forwarding the message downstream.



        LWS  =  [CRLF] 1*( SP | HT ) ; linear whitespace


   The TEXT-UTF8 rule is only used for descriptive field contents and
   values that are not intended to be interpreted by the message parser.
   Words of *TEXT-UTF8 contain characters from the UTF-8 character set
   (RFC 2279 [25]).  The TEXT-UTF8-TRIM rule is used for descriptive
   field contents that are not quoted strings, where leading and
   trailing LWS is not meaningful. In this regard, SIP differs from
   HTTP, which uses the ISO 8859-1 character set.



        TEXT-UTF8       =  *(TEXT-UTF8char | LWS)
        TEXT-UTF8-TRIM  =  *TEXT-UTF8char *(*LWS TEXT-UTF8char)
        TEXT-UTF8char   =  %x21-7e
                        |  UTF8-NONASCII
        UTF8-NONASCII   =  %xc0-df 1UTF8-CONT
                        |  %xe0-ef 2UTF8-CONT
                        |  %xf0-f7 3UTF8-CONT
                        |  %xf8-fb 4UTF8-CONT
                        |  %xfc-fd 5UTF8-CONT
        UTF8-CONT       =  %x80-bf


   A CRLF is allowed in the definition of TEXT-UTF8 only as part of a
   header field continuation. It is expected that the folding LWS will
   be replaced with a single SP before interpretation of the TEXT-UTF8
   value.

   Hexadecimal numeric characters are used in several protocol elements.



        HEX  =  "A" | "B" | "C" | "D" | "E" | "F"
                | "a" | "b" | "c" | "d" | "e" | "f" | DIGIT



Handley/Schulzrinne/Schooler/Rosenberg                      [Page 155]

Internet Draft                    SIP                      July 13, 2000


   Many SIP header field values consist of words separated by LWS or
   special characters. These special characters MUST be in a quoted
   string to be used within a parameter value.



        token                            =  1*(alphanum | "-" | "." | "!" | "%" | "*"
        | "_" | "+" | "`" | "'" | "~" )
        separators                       =  "(" | ")" | "<" | ">" | "@" |
                                            "," | ";" | ":" | "\" | <"> |
                                            "/" | "[" | "]" | "?" | "=" |
                                            "{" | "}" | SP | HT


   Comments can be included in some SIP header fields by surrounding the
   comment text with parentheses. Comments are only allowed in fields
   containing "comment" as part of their field value definition. In all
   other fields, parentheses are considered part of the field value.



        comment  =  "(" *(ctext | quoted-pair | comment) ")"
        ctext    =  < any TEXT-UTF8  excluding "("  and ")">


   A string of text is parsed as a single word if it is quoted using
   double-quote marks. In quoted strings, quotation marks (") and
   backslashes (
   ) need to be escaped.



        quoted-string  =  ( <"> *(qdtext | quoted-pair ) <"> )
        qdtext         =  LWS | %x21 | %x23-5b | %x5d-7e
                       |  UTF8-NONASCII


   The backslash character ("\") MAY be used as a single-character
   quoting mechanism only within quoted-string and comment constructs.



        quoted-pair  =  " \ " CHAR


D IANA Considerations

   Section 4.4 describes a name space and mechanism for registering SIP



Handley/Schulzrinne/Schooler/Rosenberg                      [Page 156]

Internet Draft                    SIP                      July 13, 2000


   options.

   Section 6.48 describes the name space for registering SIP warn-codes.

E Changes from RFC 2543

   In addition to editorial clarifications, this document changes or
   adds the following features to SIP as specified in RFC 2543:

        o The Record-Route header field needs to be added to each
          request and applies only to the next request.

        o Extensions developed by the IETF no longer use the org.ietf
          prefix.

        o tag syntax

        o A new optional header field, In-Reply-To (Section 6.27.

F Changes Made in Version 00

        o In Sec. 10.5.1, indicated that UAC should send both CANCEL and
          BYE after a retransmission fails.

        o Added semicolon and question mark to the list of unreserved
          characters for the user part of SIP URLs to handle tel: URLs
          properly.

        o Uniform handling of if hop count Max-Forwards: return 483.
          Note that this differs from HTTP/1.1 behavior, where only
          OPTIONS and TRACE allow this header, but respond as the final
          recipient when the value reaches zero.

        o Clarified that a forking proxy sends ACKs only for INVITE
          requests.

        o Clarified wording of DNS caching. Added paragraph on "negative
          caching", i.e., what to do if one of the hosts failed. It is
          probably not a good idea to simply drop this host from the
          list if the DNS ttl value is more than a few minutes, since
          that would mean that load balancing may not work for quite a
          while after a server is brought back on line. This will be
          true in particular if a server group receives a large number
          of requests from a small number of upstream servers, as is
          likely to be the case for calls between major consumer ISPs.
          However, without getting into arbitrary and complicated retry
          rules, it seems hard to specify any general algorithm. Might
          it be worthwhile to simply limit the "black list" interval to



Handley/Schulzrinne/Schooler/Rosenberg                      [Page 157]

Internet Draft                    SIP                      July 13, 2000


          a few minutes?

        o Added optional From-Info and Alert-Info header fields that
          describe the caller and information to be used in alerting.
          (Currently, avoided use of "purpose" qualification since it is
          not yet clear whether rendering content without understanding
          its meaning is always appropriate. For example, if a UAS does
          not understand that this header is to replace ringing, it
          would mix both local ring tone and the indicated sound URL.)
          TBD!

        o SDP "s=" lines can't be empty, unfortunately.  (Section B.7)

        o Noted that maddr could also contain a unicast address, but
          SHOULD contain the multicast address if the request is sent
          via multicast (Section 6.47, 10.2.2).

        o Clarified that responses are sent to port in Via sent-by
          value.

        o Added "other-*" to the user URL parameter and the Hide and
          Content-Disposition headers.

        o Clarify generation of timeout (408) responses in forking
          proxies and mention the Expires header.  (Section 12.4)

        o Clarified that CANCEL and INVITE are separate transactions
          (Fig. 12). Thus, the INVITE request generates a 487
          (Transaction Terminated) if a CANCEL or BYE arrives.

        o Clarified that Record-Route SHOULD be inserted in every
          request, but that the route, once established, persists. This
          provides robustness if the called UAS crashes.

        o Emphasized that proxy, redirect, registrar and location
          servers are logical, not physical entities and that UAC and
          UAS roles are defined on a request-by-request basis. (Section
          1.3)

        o In Section 6.47, noted that the maddr and received parameters
          also need to be encrypted when doing Via hiding.

        o Simplified Fig. 12 to only show INVITE transaction.

        o Added definition of the use of Contact (Section 6.15) for
          OPTIONS.

        o Added HTTP/RFC822 headers Content-Language and MIME-Version.



Handley/Schulzrinne/Schooler/Rosenberg                      [Page 158]

Internet Draft                    SIP                      July 13, 2000


        o Added note in Section A indicating that UAs need to support
          UDP.

        o Added explanation in Section 11.5 explaining what a UA should
          do when receiving an initial INVITE with a tag.

        o Clarified UA and proxy behavior for 302 responses (Section
          7.3.3).

        o Added details on what a UAS should do when receiving a tagged
          INVITE request for an unknown call leg. This could occur if
          the UAS had crashed and the UAC sends a re-INVITE or if the
          BYE got lost and the UAC still believes to be in the call.

        o Added definition of Contact in 4xx, 5xx and 6xx to "redirect"
          to more error details.

        o Added note to forking proxy description in Section 12.4 to
          gather *-Authenticate from responses. This allows several
          branches to be authenticated simultaneously.

        o Changed URI syntax to use URL escaping instead of quotation
          marks.

        o Changed SIP URL definition to reference RFC 2806 for
          telephone-subscriber part.

        o Clarify that the To URI should basically be ignored by the
          receiving UAS except for matching requests to call legs. In
          particular, To headers with a scheme or name unknown to the
          callee should be accepted.

        o Clarify in Section 6.47.1 that maddr is to be added by any
          client, either proxy or UAC.

        o Added response code 488 to indicate that there was no common
          media at the particular destination. (606 indicates such
          failure globally.)

        o In Section 6.24, noted that registration updates can shorten
          the validity period.

        o Added note to Section 14.3 to enclose the URI in quotation
          marks. The BNF in RFC 2617 is in error.

        o Clarified that registrars use Authorization and WWW-
          Authenticate, not proxy authentication.




Handley/Schulzrinne/Schooler/Rosenberg                      [Page 159]

Internet Draft                    SIP                      July 13, 2000


        o Added note in Section 6.15 that "headers" are copied from
          Contact into the new request.

        o Changed URL syntax so that port specifications have to have at
          least one digit, in line with other URL formats such as
          "http".  Previously, an empty port number was permissible.

        o In Section B, added a section on how to add and delete streams
          in re-INVITEs.

        o IETF-blessed extensions now have short names, without
          org.ietf. prefix.

        o Cseq is unique within a call leg, not just within a call
          (Section 6.21).

        o Added IPv6 literal addresses to the SIP URL definition in
          Section 2, according to RFC 2732 [48].  Modified the IPv4
          address to limit segments to at most three digits.

        o In Section 4.2.6, modify registration procedure so that it
          explicitly references the URL comparison. Updates with shorter
          expiration time are now allowed.

        o For send-only media, SDP still must indicate the address and
          port, since these are needed as destinations for RTCP
          messages.  (Section B)

        o Changed references regarding DNS SRV records from RFC 2052 to
          RFC 2782, which is now a Proposed Standard. Integrated SRV
          into the search procedure in Section 1 and removed the SRV
          appendix. The only visible change is that protocol and service
          names are now prefixed by an underscore. Added wording that
          incorporates the precedence of maddr.

        o Allow parameters in Record-Route and Route headers.

        o In Table 2, list udp as the default value for the transport
          parameter in SIP URI.

        o Removed sentence that From can be encrypted. It cannot, since
          the header is needed for call-leg identification.

        o Added note that a UAC only copies a To tag into subsequent
          transactions if it arrives in a 200 OK to an INVITE in Section
          11. This avoids the problem that occurs when requests get
          resubmitted after receiving, say, a 407 (or possibly 500, 503,
          504, 305, 400, 411, 413, maybe even 408). Under the old rules,



Handley/Schulzrinne/Schooler/Rosenberg                      [Page 160]

Internet Draft                    SIP                      July 13, 2000


          these requests would have a tag, which would force the called
          UAS to reject the request, since it doesn't have an entry for
          this tag.

        o Loop detection has been modified to take the request-URI into
          account (Section 12.3 and 6.47.5). This allows the same
          request to visit the server twice, but with different request
          URIs.

        o Elaborated on URL comparison and comparison of From/To fields.

        o Added np-queried user parameter.

        o Changed tag syntax from UUID to token, since there's no reason
          to restrict it to hex.

        o Added Content-Disposition header based on earlier discussions
          about labeling what to do with a message body (part).

        o Clarification: proxies must insert To tags for locally
          generated responses.

        o Clarification: multicast may be used for subsequent
          registrations.

        o Feature: Added Supported header. Needed if client wants to
          indicate things the server can usefully return in the
          response.

        o Bug: The From, To, and Via headers were missing extension
          parameters. The Encryption and Response-Key header fields now
          "officially" allow parameters consisting only of a token,
          rather than just "token = value".

        o Bug: Allow was listed as optional in 405 responses in Table 4.
          It is mandatory.

        o Added in Section 4.2.4: "A BYE request from either called or
          calling party terminates any pending INVITE, but the INVITE
          request transaction MUST be completed with a final response."

        o Clarified in Section 4.2.1: "If an INVITE request for an
          existing session fails, the session description agreed upon in
          the last successful INVITE transaction remains in force."

        o Clarified in Section 4.2.1 what happens if two INVITE requests
          meet each other on the wire, either traveling the same or in
          opposite directions:



Handley/Schulzrinne/Schooler/Rosenberg                      [Page 161]

Internet Draft                    SIP                      July 13, 2000


             A UAC MUST NOT issue another INVITE request for the
             same call leg before the previous transaction has
             completed. A UAS that receives an INVITE before it
             sent the final response to an INVITE with a lower CSeq
             number MUST return a 400 (Bad Request) response and
             MUST include a Retry-After header field with a
             randomly chosen value of between 0 and 10 seconds. A
             UA that receives an INVITE while it has an INVITE
             transaction pending, returns a 500 (Internal Server
             Error) and also includes a Retry-After header field.

        o Expires header clarified: limits only duration of INVITE
          transaction, not the actual session. SDP does the latter.

        o The In-Reply-To header was added (Section 6.27).

        o There were two incompatible BNFs for WWW-Authenticate.  One
          defined for PGP, and the other borrowed from HTTP. For basic
          or digest:


            WWW-Authenticate: basic realm="Wallyworld"



          and for pgp:


            WWW-Authenticate: pgp; realm="Wallyworld"



          The latter is incorrect and the semicolon has been removed.

        o Added rules for Route construction from called to calling UA.

        o We now allow Accept and Accept-Encoding in BYE and CANCEL
          requests. There is no particular reason not to allow them, as
          both requests could theoretically return responses,
          particularly when interworking with other signaling systems.

        o PGP "pgp-pubalgorithm" allows server to request the desired
          public-key algorithm.

        o ABNF rules now describe tokens explicitly rather than by
          subtraction; explicit character enumeration for CTL, etc.

        o Registrars should be careful to check the Date header as the



Handley/Schulzrinne/Schooler/Rosenberg                      [Page 162]

Internet Draft                    SIP                      July 13, 2000


          expiration time may well be in the past, as seen by the
          client.

        o Content-Length is mandatory; Table 4 erroneously marked it as
          optional.

        o User-Agent was classified in a syntax definition as a request
          header rather than a general header.

        o Clarify ordering of items to be signed and include realm in
          list.

        o Allow Record-Route in 401 and 484 responses.

        o Hop-by-hop need to preceded end-to-end headers only if
          authentication is used (Section 6).

        o 1xx message bodies MAY now contain session descriptions.

        o Changed references to HTTP/1.1 and authentication to point to
          the latest RFCs.

        o Added 487 (Request terminated) status response. It is issued
          if the original request was terminated via CANCEL or BYE.

        o The spec was not clear on the identification of a call leg.
          Section 1.3 says it's the combination of To, From, and Call-
          ID. However, requests from the callee to the caller have the
          To and From reversed, so this definition is not quite
          accurate. Additionally, the "tag" field should be included in
          the definition of call leg. The spec now says that a call leg
          is defined as the combination of local-address, remote-
          address, and call-id, where these addresses include tags.

          Text was added to Section 6.21 to emphasize that the From and
          To headers designate the originator of the request, not that
          of the call leg.

G Changes To Be Made

   NOTE: Almost all of these require further discussion or may already
   have been integrated into the main spec.

        o Bug: All URI parameters, except method, are allowed in a
          Request-URI. Consequently, also updated the description of
          which parameters are copied from 3xx responses in Sec. 6.15.

        o Bug: Consider a proxy that implements an ACD service. It



Handley/Schulzrinne/Schooler/Rosenberg                      [Page 163]

Internet Draft                    SIP                      July 13, 2000


          proxies call requests to operators based on some complex
          logic. After receiving and INVITE, proxying it, and forwarding
          the response, the proxy eventually times out its state. Later
          on, an ACK (a late one, or one that has been wandering in the
          network) arrives. Where should the proxy send this? Ideally,
          the proxy should know where it sent the INVITE, and send it to
          the same place. This is not workable for complex forwarding
          logic. To fix this, the ACKs should go directly to the UAS.
          This will occur if a UAS inserts a Contact into the 200 OK,
          and if the UAC honors this in subsequent requests (including
          ACK). However, insertion of Contact, and honoring of Contact,
          are SHOULD, not MUST, so this "lost ACK" case is still
          possible in conditionally compliant implementations. TO fix
          this, (1) insertion of a Contact header, and (2) honoring of a
          Contact header, should be made MUST strength. (Q: is this the
          only solution???)

        o Bug: The use of CRLF,CR,or LF to terminate lines is confusing.
          Basically, each header line can be terminated by a CR, LF, or
          CRLF.  Furthermore, the end of the headers is signified by a
          "double return".  Thus, if the BNF is defined as:

          return = (CR | LF | CRLF)

          and double-return is (CRLF CRLF)

          This grammar is not context-free, since CR LF could be either
          return or double return, and thus you may not be able to
          distinguish the end of line with end of headers. In fact, the
          header section should be terminated by a double-return of the
          form:  double-return = (CR CR) | (LF LF) | (CR LF CR LF)

          This is inconsistent with the BNF currently defined in the
          spec, which reads like the former definition. This needs to be
          explicitly clarified in the spec.

        o Bug: Section 4.2.6 specifies that, for a REGISTER request, the
          user name of the Request-URI MUST be empty. While it does not
          mention anything about the "@" sign, the SIP-URL BNF is
          defined so that an "@" sign is used only in conjunction with a
          user ID. The implication that no "@" sign appears in a
          REGISTER Request-URI is never explicitly stated.

        o Bug: Round brackets in Contact header are sort of legacy, and
          very hard to implement. They are also not that useful. Perhaps
          they should be removed. Vote during bakeoff was that this is
          OK.




Handley/Schulzrinne/Schooler/Rosenberg                      [Page 164]

Internet Draft                    SIP                      July 13, 2000


        o Bug: The spec says that a proxy is a back to back UAS/UAC.
          This is almost, but not quite, true. For example, a UAS should
          insert a tag into a provisional response, but a proxy should
          not. This should be clarified.

        o Issue: Usage of Content-Length. Spec is not that clear under
          what conditions its needed, particularly with TCP. This should
          be clarified. Another issue: should Content-Length just be
          made mandatory? It's not as a carryover from HTTP; do we have
          the same assumptions they do? Also, the section on Content-
          Length does not discuss behavior for requests over TCP that do
          not contain Content-Lengths (Answer: if no Content-Length is
          present in requests, the request packet must be terminated
          with a close connection. However, having the client close a
          connection is the same as sending a CANCEL, so no response is
          ever received. Now, a client *can* do this, but it makes no
          sense)

        o Issue: In REGISTER requests, should the request URI be a fully
          qualified hostname, or the domainname:


            REGISTER sip:machine.company.com SIP/2.0



          or

            REGISTER sip:company.com SIP/2.0



          It matters since registrars check if the domain name is
          there's before accepting the registration. The comparison
          might fail if the wrong thing is sent. Answer: probably a
          proxy/UAC SHOULD send it to company.com (and then use SRV to
          find the right machine), but a registrar SHOULD accept both
          fully qualified host names and domain names in comparisons. In
          fact, the registrar probably should be making this comparison
          on the To field, not the Request-URI!

        o Issue: A sends an INVITE to B, and B returns a 180 with a tag.
          Then, A sends BYE to B, without a tag in the To field. Should
          B accept this BYE or not? Answer: not sure; I think actually
          it says somewhere in the spec you are not supposed to initiate
          a new transaction until the previous completes.

        o Bug: No default value for the q parameter in Contact is



Handley/Schulzrinne/Schooler/Rosenberg                      [Page 165]

Internet Draft                    SIP                      July 13, 2000


          defined.  This is not strictly needed, but is useful for
          consistent behaviors at recursive proxies and at UAC's

        o Bug: A INVITEs B through a proxy, B responds to A with a tag
          in the To field. Later, A sends a re-INVITE to B, with a tag
          in the To field, through the proxy. This request is malformed,
          so the proxy returns a 400 error. Normally, a proxy will
          insert a tag into the To field of a 400 response. But here, it
          cannot, since a tag is already present in the request. Thus,
          the rule should be that if a request arrives with a tag, the
          entity sending a response MUST NOT insert an additional tag,
          even if it a different entity than the one that created the
          tag in the first place.

        o Bug: When UAS mirrors the To field in the response, and the
          UAC "matches" it to determine what call and call-leg it is, it
          MUST make the comparison based on the user, host, and tag of
          the URI only - not based on the display name, existence of
          brackets, etc. (Basically, don't do a pure string match).

        o Bug: When multicasting a register, if a server wishes to
          authenticate the user it, must return a 401. However, it is
          not permitted to send a 400 response via multicast. How should
          this be handled?

        o Bug: If a proxy sends a request by UDP (TCP), the spec does
          not disallow placing TCP (UDP) in the transport parameter of
          the Via field, which it should.

        o Bug: The spec says that if a server receives a CANCEL for
          which it has no pending transaction, it returns an error.
          This, however, is not true for a proxy. A stateless proxy
          SHOULD forward the CANCEL as it would any non-INVITE request.

        o Bug: Currently, the spec says a proxy should not insert a
          Record-Route into a request with Route headers. I think this
          is wrong. The Route should be refreshed with each new request
          by a new Record-Route. The primary motivation here is
          mobility; if a user moves and sends a new request with a new
          Contact header, the Route must change. Similarly, this new
          route may require new proxies on the route (or old ones to
          drop off).

        o Section 6.13 in the rfc begins mid-paragraph after the BNF.
          The following text seems to have been misplaced in the
          conversion to ASCII:

             Even if the "display-name" is empty, the "name-addr"



Handley/Schulzrinne/Schooler/Rosenberg                      [Page 166]

Internet Draft                    SIP                      July 13, 2000


             form MUST be used if the "addr-spec" contains a comma,
             semicolon or question mark.

        o Table 5 lists WWW-Authenticate as a response-header allowed
          only in 401 responses. However, section 13.2 allows its reuse
          in requests containing Require: signed-response header. Also
          examples in section 14.1 have WWW-Authenticate in INVITE
          requests with no Require: present.

          A client MAY require that a server sign its response by
          including a Require: signed-response request header field.
          The client indicates the desired authentication method via the
          WWW- Authenticate header.

          This is also an inconsistency. If we do allow this, then WWW-
          Authenticate must be a request and response header field, and
          the Authorization header as well, since then it is included in
          a signed response.

        o Text on including Authorization in responses and WWW-
          Authenticate in requests needs to be added.

        o Ping liveness detection. Motivation:

             A proxy may be keeping track of a call during its
             entire duration, for example to maintain consistent
             routing for all subsequent requests. The proxy
             enforces that requests traverse it by using Record-
             Route. If, however, a BYE fails to arrive at the
             termination of a call (e.g., because of application
             failure or network partition), the proxy must know to
             delete the call record. If the proxy deletes a call
             record before a session is terminated (by timeout, for
             example), it makes the routing of subsequent call
             messages impossible (since the information regarding
             the ultimate destination of an INVITE may no longer be
             available). Without the ability for a proxy server to
             check the liveness of a session, there is no way to
             rectify this problem.

          Solutions suggested:

        o proxy sends re-INVITE;

        o session lifetime;

        o A PING request would be retransmitted (T1, doubled each time,
          max of T2, up to 11 times) until a provisional or final



Handley/Schulzrinne/Schooler/Rosenberg                      [Page 167]

Internet Draft                    SIP                      July 13, 2000


          response is received.  Once that is accomplished, it is only
          retransmitted at the frequency specified in the "Period"
          header. As long as another provisional response is received in
          at the freqency specified in the "Period" header, the PING is
          not retransmitted.

        o The OPTIONS request is answered with a special 2xx status if
          the call leg exists at the server. (What about changes in
          SDP?)

        o received-port for NAPT support in Via header field;

        o Proxy authentication: If a forking proxy adds or changes an
          Authorization header, it MUST change the branch parameter in
          the Via header to avoid being confused by retransmission of
          earlier 407 responses. If a request triggered a 407 response,
          the downstream proxy inspects retransmissions of the the same
          request to see if it contains new authorization header. If so,
          it re-evaluates its authentication decision. This is a slight
          deviation from the usual idempotency rules.

H Acknowledgments

   We wish to thank the members of the IETF MMUSIC and SIP WGs for their
   comments and suggestions. Detailed comments were provided by Jim
   Buller, Neil Deason, Dave Devanathan, Cédric Fluckiger, Yaron Goland,
   Bernie Höneisen, Phil Hoffer, Christian Huitema, Jean Jervis, Gadi
   Karmi, Anders Kristensen, Jonathan Lennox, Gethin Liddell, Keith
   Moore, Vern Paxson, Moshe J. Sambol, Chip Sharp, Igor Slepchin,
   Robert Sparks, Eric Tremblay., and Rick Workman.

   This work is based, inter alia, on [49,50].

I Authors' Addresses

   Mark Handley
   ACIRI
   electronic mail:  mjh@aciri.org

   Henning Schulzrinne
   Dept. of Computer Science
   Columbia University
   1214 Amsterdam Avenue
   New York, NY 10027
   USA
   electronic mail:  schulzrinne@cs.columbia.edu

   Eve Schooler



Handley/Schulzrinne/Schooler/Rosenberg                      [Page 168]

Internet Draft                    SIP                      July 13, 2000


   Computer Science Department 256-80
   California Institute of Technology
   Pasadena, CA 91125
   USA
   electronic mail:  schooler@cs.caltech.edu

   Jonathan Rosenberg
   dynamicsoft
   72 Eagle Rock Ave
   East Hanover, NJ 07936
   USA
   electronic mail:  jdrosen@dynamicsoft.com

J Bibliography

   [1] R. Pandya, "Emerging mobile and personal communication systems,"
   IEEE Communications Magazine , Vol. 33, pp. 44--52, June 1995.

   [2] R. Braden, Ed., L. Zhang, S. Berson, S. Herzog, and S. Jamin,
   "Resource ReSerVation protocol (RSVP) -- version 1 functional
   specification," Request for Comments 2205, Internet Engineering Task
   Force, Sept. 1997.

   [3] H. Schulzrinne, S. Casner, R. Frederick, and V. Jacobson, "RTP: a
   transport protocol for real-time applications," Request for Comments
   1889, Internet Engineering Task Force, Jan. 1996.

   [4] H. Schulzrinne, A. Rao, and R. Lanphier, "Real time streaming
   protocol (RTSP)," Request for Comments 2326, Internet Engineering
   Task Force, Apr.  1998.

   [5] M. Handley, "SAP: Session announcement protocol," Internet Draft,
   Internet Engineering Task Force, Nov. 1996.  Work in progress.

   [6] M. Handley and V. Jacobson, "SDP: session description protocol,"
   Request for Comments 2327, Internet Engineering Task Force, Apr.
   1998.

   [7] International Telecommunication Union, "Visual telephone systems
   and equipment for local area networks which provide a non-guaranteed
   quality of service," Recommendation H.323, Telecommunication
   Standardization Sector of ITU, Geneva, Switzerland, May 1996.

   [8] International Telecommunication Union, "Control protocol for
   multimedia communication," Recommendation H.245, Telecommunication
   Standardization Sector of ITU, Geneva, Switzerland, Feb. 1998.

   [9] International Telecommunication Union, "Media stream



Handley/Schulzrinne/Schooler/Rosenberg                      [Page 169]

Internet Draft                    SIP                      July 13, 2000


   packetization and synchronization on non-guaranteed quality of
   service LANs," Recommendation H.225.0, Telecommunication
   Standardization Sector of ITU, Geneva, Switzerland, Nov. 1996.

   [10] S. Bradner, "Key words for use in RFCs to indicate requirement
   levels," Request for Comments 2119, Internet Engineering Task Force,
   Mar. 1997.

   [11] R. Fielding, J. Gettys, J. Mogul, H. Frystyk, L. Masinter, P.
   Leach, and T. Berners-Lee, "Hypertext transfer protocol -- HTTP/1.1,"
   Request for Comments 2616, Internet Engineering Task Force, June
   1999.

   [12] T. Berners-Lee, R. Fielding, and L. Masinter, "Uniform resource
   identifiers (URI): generic syntax," Request for Comments 2396,
   Internet Engineering Task Force, Aug. 1998.

   [13] T. Berners-Lee, L. Masinter, and M. McCahill, "Uniform resource
   locators (URL)," Request for Comments 1738, Internet Engineering Task
   Force, Dec.  1994.

   [14] A. Gulbrandsen, P. Vixie, and L. Esibov, "A DNS RR for
   specifying the location of services (DNS SRV)," Request for Comments
   2782, Internet Engineering Task Force, Feb. 2000.

   [15] M. Hamilton and R. Wright, "Use of DNS aliases for network
   services," Request for Comments 2219, Internet Engineering Task
   Force, Oct. 1997.

   [16] P. V. Mockapetris, "Domain names - implementation and
   specification," Request for Comments 1035, Internet Engineering Task
   Force, Nov. 1987.

   [17] D. Zimmerman, "The finger user information protocol," Request
   for Comments 1288, Internet Engineering Task Force, Dec. 1991.

   [18] S. Williamson, M. Kosters, D. Blacka, J. Singh, and K. Zeilstra,
   "Referral whois (rwhois) protocol V1.5," Request for Comments 2167,
   Internet Engineering Task Force, June 1997.

   [19] W. Yeong, T. Howes, and S. Kille, "Lightweight directory access
   protocol," Request for Comments 1777, Internet Engineering Task
   Force, Mar. 1995.

   [20] E. M. Schooler, "A multicast user directory service for
   synchronous rendezvous," Master's Thesis CS-TR-96-18, Department of
   Computer Science, California Institute of Technology, Pasadena,
   California, Aug. 1996.



Handley/Schulzrinne/Schooler/Rosenberg                      [Page 170]

Internet Draft                    SIP                      July 13, 2000


   [21] A. Vaha-Sipila, "URLs for telephone calls," Request for Comments
   2806, Internet Engineering Task Force, Apr. 2000.

   [22] J. Postel, "User datagram protocol," Request for Comments 768,
   Internet Engineering Task Force, Aug. 1980.

   [23] J. Postel, "DoD standard transmission control protocol," Request
   for Comments 761, Internet Engineering Task Force, Jan. 1980.

   [24] T. Dierks and C. Allen, "The TLS protocol version 1.0," Request
   for Comments 2246, Internet Engineering Task Force, Jan. 1999.

   [25] F. Yergeau, "UTF-8, a transformation format of ISO 10646,"
   Request for Comments 2279, Internet Engineering Task Force, Jan.
   1998.

   [26] W. R. Stevens, TCP/IP illustrated: the protocols , Vol. 1.
   Reading, Massachusetts: Addison-Wesley, 1994.

   [27] J. C. Mogul and S. E. Deering, "Path MTU discovery," Request for
   Comments 1191, Internet Engineering Task Force, Nov. 1990.

   [28] D. Crocker, "Standard for the format of ARPA internet text
   messages," Request for Comments 822, Internet Engineering Task Force,
   Aug. 1982.

   [29] D. Meyer, "Administratively scoped IP multicast," Request for
   Comments 2365, Internet Engineering Task Force, July 1998.

   [30] H. Schulzrinne, "RTP profile for audio and video conferences
   with minimal control," Request for Comments 1890, Internet
   Engineering Task Force, Jan.  1996.

   [31] D. Eastlake, 3rd, S. Crocker, and J. Schiller, "Randomness
   recommendations for security," Request for Comments 1750, Internet
   Engineering Task Force, Dec.  1994.

   [32] F. Dawson and T. Howes, "vcard MIME directory profile," Request
   for Comments 2426, Internet Engineering Task Force, Sept. 1998.

   [33] G. Good, "The LDAP data interchange format (LDIF) - technical
   specification," Request for Comments 2849, Internet Engineering Task
   Force, June 2000.

   [34] P. Hoffman, L. Masinter, and J. Zawinski, "The mailto URL
   scheme," Request for Comments 2368, Internet Engineering Task Force,
   July 1998.




Handley/Schulzrinne/Schooler/Rosenberg                      [Page 171]

Internet Draft                    SIP                      July 13, 2000


   [35] R. Troost and S. Dorner, "Communicating presentation information
   in internet messages: The content-disposition header," Request for
   Comments 1806, Internet Engineering Task Force, June 1995.

   [36] R. T. Braden, "Requirements for internet hosts - application and
   support," Request for Comments 1123, Internet Engineering Task Force,
   Oct. 1989.

   [37] J. Palme, "Common internet message headers," Request for
   Comments 2076, Internet Engineering Task Force, Feb. 1997.

   [38] R. Rivest, "The MD5 message-digest algorithm," Request for
   Comments 1321, Internet Engineering Task Force, Apr. 1992.

   [39] H. Alvestrand, "IETF policy on character sets and languages,"
   Request for Comments 2277, Internet Engineering Task Force, Jan.
   1998.

   [40] M. Elkins, "MIME security with pretty good privacy (PGP),"
   Request for Comments 2015, Internet Engineering Task Force, Oct.
   1996.

   [41] N. Freed and N. Borenstein, "Multipurpose internet mail
   extensions (MIME) part two: Media types," Request for Comments 2046,
   Internet Engineering Task Force, Nov. 1996.

   [42] D. Atkins, W. Stallings, and P. Zimmermann, "PGP message
   exchange formats," Request for Comments 1991, Internet Engineering
   Task Force, Aug. 1996.

   [43] R. Atkinson, "Security architecture for the internet protocol,"
   Request for Comments 1825, Internet Engineering Task Force, Aug.
   1995.

   [44] J. Franks, P. Hallam-Baker, J. Hostetler, S. Lawrence, P. Leach,
   A. Luotonen, and L. Stewart, "HTTP authentication: Basic and digest
   access authentication," Request for Comments 2617, Internet
   Engineering Task Force, June 1999.

   [45] J. Callas, L. Donnerhacke, H. Finney, and R. Thayer, "OpenPGP
   message format," Request for Comments 2440, Internet Engineering Task
   Force, Nov.  1998.

   [46] N. Freed and N. Borenstein, "Multipurpose internet mail
   extensions (MIME) part one: Format of internet message bodies,"
   Request for Comments 2045, Internet Engineering Task Force, Nov.
   1996.




Handley/Schulzrinne/Schooler/Rosenberg                      [Page 172]

Internet Draft                    SIP                      July 13, 2000


   [47] D. Crocker, Ed., and P. Overell, "Augmented BNF for syntax
   specifications:  ABNF," Request for Comments 2234, Internet
   Engineering Task Force, Nov.  1997.

   [48] R. Hinden, B. Carpenter, and L. Masinter, "Format for literal
   IPv6 addresses in URL's," Request for Comments 2732, Internet
   Engineering Task Force, Dec. 1999.

   [49] E. M. Schooler, "Case study: multimedia conference control in a
   packet-switched teleconferencing system," Journal of Internetworking:
   Research and Experience , Vol. 4, pp. 99--120, June 1993.  ISI
   reprint series ISI/RS-93-359.

   [50] H. Schulzrinne, "Personal mobility for multimedia services in
   the Internet," in European Workshop on Interactive Distributed
   Multimedia Systems and Services (IDMS) , (Berlin, Germany), Mar.
   1996.


   Full Copyright Statement

   Copyright (c) The Internet Society (2000). All Rights Reserved.

   This document and translations of it may be copied and furnished to
   others, and derivative works that comment on or otherwise explain it
   or assist in its implementation may be prepared, copied, published
   and distributed, in whole or in part, without restriction of any
   kind, provided that the above copyright notice and this paragraph are
   included on all such copies and derivative works. However, this
   document itself may not be modified in any way, such as by removing
   the copyright notice or references to the Internet Society or other
   Internet organizations, except as needed for the purpose of
   developing Internet standards in which case the procedures for
   copyrights defined in the Internet Standards process must be
   followed, or as required to translate it into languages other than
   English.

   The limited permissions granted above are perpetual and will not be
   revoked by the Internet Society or its successors or assigns.

   This document and the information contained herein is provided on an
   "AS IS" basis and THE INTERNET SOCIETY AND THE INTERNET ENGINEERING
   TASK FORCE DISCLAIMS ALL WARRANTIES, EXPRESS OR IMPLIED, INCLUDING
   BUT NOT LIMITED TO ANY WARRANTY THAT THE USE OF THE INFORMATION
   HEREIN WILL NOT INFRINGE ANY RIGHTS OR ANY IMPLIED WARRANTIES OF
   MERCHANTABILITY OR FITNESS FOR A PARTICULAR PURPOSE.





Handley/Schulzrinne/Schooler/Rosenberg                      [Page 173]

Internet Draft                    SIP                      July 13, 2000


                           Table of Contents



   1          Introduction ........................................    1
   1.1        Overview of SIP Functionality .......................    1
   1.2        Terminology .........................................    3
   1.3        Definitions .........................................    3
   1.4        Overview of SIP Operation ...........................    7
   1.4.1      SIP Addressing ......................................    8
   1.4.2      Locating a SIP Server ...............................    8
   1.4.3      SIP Transaction .....................................   11
   1.4.4      SIP Invitation ......................................   11
   1.4.5      Locating a User .....................................   13
   1.4.6      Changing an Existing Session ........................   15
   1.4.7      Registration Services ...............................   15
   1.5        Protocol Properties .................................   16
   1.5.1      Minimal State .......................................   16
   1.5.2      Lower-Layer-Protocol Neutral ........................   16
   1.5.3      Text-Based ..........................................   16
   2          SIP Uniform Resource Locators .......................   16
   2.1        SIP URL Comparison ..................................   21
   2.2        Non-SIP URLs ........................................   22
   3          SIP Message Overview ................................   22
   4          Request .............................................   24
   4.1        Request-Line ........................................   24
   4.2        Methods .............................................   24
   4.2.1      INVITE ..............................................   26
   4.2.2      ACK .................................................   28
   4.2.3      OPTIONS .............................................   29
   4.2.4      BYE .................................................   29
   4.2.5      CANCEL ..............................................   30
   4.2.6      REGISTER ............................................   31
   4.3        Request-URI .........................................   35
   4.3.1      SIP Version .........................................   36
   4.4        Option Tags .........................................   36
   4.4.1      Registering New Option Tags with IANA ...............   36
   5          Response ............................................   37
   5.1        Status-Line .........................................   37
   5.1.1      Status Codes and Reason Phrases .....................   38
   6          Header Field Definitions ............................   40
   6.1        General Header Fields ...............................   42
   6.2        Entity Header Fields ................................   42
   6.3        Request Header Fields ...............................   43
   6.4        Response Header Fields ..............................   44
   6.5        End-to-end and Hop-by-hop Headers ...................   45
   6.6        Header Field Format .................................   45
   6.7        Accept ..............................................   46



Handley/Schulzrinne/Schooler/Rosenberg                      [Page 174]

Internet Draft                    SIP                      July 13, 2000


   6.8        Accept-Encoding .....................................   46
   6.9        Accept-Language .....................................   47
   6.10       Alert-Info ..........................................   47
   6.11       Allow ...............................................   47
   6.12       Authorization .......................................   48
   6.13       Call-ID .............................................   48
   6.14       Call-Info ...........................................   50
   6.15       Contact .............................................   50
   6.16       Content-Disposition .................................   54
   6.17       Content-Encoding ....................................   54
   6.18       Content-Language ....................................   55
   6.19       Content-Length ......................................   55
   6.20       Content-Type ........................................   56
   6.21       CSeq ................................................   56
   6.22       Date ................................................   58
   6.23       Encryption ..........................................   59
   6.24       Expires .............................................   60
   6.25       From ................................................   61
   6.26       Hide ................................................   63
   6.27       In-Reply-To .........................................   64
   6.28       Max-Forwards ........................................   64
   6.29       MIME-Version ........................................   65
   6.30       Organization ........................................   65
   6.31       Priority ............................................   65
   6.32       Proxy-Authenticate ..................................   66
   6.33       Proxy-Authorization .................................   67
   6.34       Proxy-Require .......................................   67
   6.35       Record-Route ........................................   67
   6.35.1     Operation ...........................................   67
   6.35.2     Construction of Route Header ........................   68
   6.35.3     Request Destination .................................   69
   6.35.4     Syntax ..............................................   69
   6.35.5     Example .............................................   69
   6.36       Require .............................................   69
   6.37       Response-Key ........................................   70
   6.38       Retry-After .........................................   71
   6.39       Route ...............................................   72
   6.40       Server ..............................................   72
   6.41       Subject .............................................   72
   6.42       Supported ...........................................   72
   6.43       Timestamp ...........................................   73
   6.44       To ..................................................   73
   6.45       Unsupported .........................................   75
   6.46       User-Agent ..........................................   75
   6.47       Via .................................................   75
   6.47.1     Requests ............................................   75
   6.47.2     Receiver-tagged Via Header Fields ...................   76
   6.47.3     Responses ...........................................   77



Handley/Schulzrinne/Schooler/Rosenberg                      [Page 175]

Internet Draft                    SIP                      July 13, 2000


   6.47.4     User Agent and Redirect Servers .....................   78
   6.47.5     Syntax ..............................................   78
   6.48       Warning .............................................   80
   6.49       WWW-Authenticate ....................................   82
   7          Status Code Definitions .............................   82
   7.1        Informational 1xx ...................................   83
   7.1.1      100 Trying ..........................................   83
   7.1.2      180 Ringing .........................................   83
   7.1.3      181 Call Is Being Forwarded .........................   83
   7.1.4      182 Queued ..........................................   83
   7.1.5      183 Session Progress ................................   83
   7.2        Successful 2xx ......................................   84
   7.2.1      200 OK ..............................................   84
   7.3        Redirection 3xx .....................................   84
   7.3.1      300 Multiple Choices ................................   84
   7.3.2      301 Moved Permanently ...............................   85
   7.3.3      302 Moved Temporarily ...............................   85
   7.3.4      305 Use Proxy .......................................   85
   7.3.5      380 Alternative Service .............................   86
   7.4        Request Failure 4xx .................................   86
   7.4.1      400 Bad Request .....................................   86
   7.4.2      401 Unauthorized ....................................   86
   7.4.3      402 Payment Required ................................   86
   7.4.4      403 Forbidden .......................................   86
   7.4.5      404 Not Found .......................................   86
   7.4.6      405 Method Not Allowed ..............................   86
   7.4.7      406 Not Acceptable ..................................   87
   7.4.8      407 Proxy Authentication Required ...................   87
   7.4.9      408 Request Timeout .................................   87
   7.4.10     409 Conflict ........................................   87
   7.4.11     410 Gone ............................................   87
   7.4.12     411 Length Required .................................   88
   7.4.13     413 Request Entity Too Large ........................   88
   7.4.14     414 Request-URI Too Long ............................   88
   7.4.15     415 Unsupported Media Type ..........................   88
   7.4.16     420 Bad Extension ...................................   88
   7.4.17     480 Temporarily Unavailable .........................   88
   7.4.18     481 Call Leg/Transaction Does Not Exist .............   89
   7.4.19     482 Loop Detected ...................................   89
   7.4.20     483 Too Many Hops ...................................   89
   7.4.21     484 Address Incomplete ..............................   89
   7.4.22     485 Ambiguous .......................................   89
   7.4.23     486 Busy Here .......................................   90
   7.4.24     487 Request Terminated ..............................   90
   7.4.25     488 Not Acceptable Here .............................   90
   7.5        Server Failure 5xx ..................................   90
   7.5.1      500 Server Internal Error ...........................   90
   7.5.2      501 Not Implemented .................................   91



Handley/Schulzrinne/Schooler/Rosenberg                      [Page 176]

Internet Draft                    SIP                      July 13, 2000


   7.5.3      502 Bad Gateway .....................................   91
   7.5.4      503 Service Unavailable .............................   91
   7.5.5      504 Server Time-out .................................   91
   7.5.6      505 Version Not Supported ...........................   91
   7.6        Global Failures 6xx .................................   92
   7.6.1      600 Busy Everywhere .................................   92
   7.6.2      603 Decline .........................................   92
   7.6.3      604 Does Not Exist Anywhere .........................   92
   7.6.4      606 Not Acceptable ..................................   92
   8          SIP Message Body ....................................   92
   8.1        Body Inclusion ......................................   93
   8.2        Message Body Type ...................................   93
   8.3        Message Body Length .................................   93
   9          Compact Form ........................................   94
   10         Behavior of SIP Clients and Servers .................   95
   10.1       General Remarks .....................................   95
   10.1.1     Requests ............................................   95
   10.1.2     Responses ...........................................   95
   10.2       Source Addresses, Destination Addresses and
   Connections ....................................................   96
   10.2.1     Unicast UDP .........................................   96
   10.2.2     Multicast UDP .......................................   96
   10.3       TCP .................................................   97
   10.4       Reliability for Requests Other Than INVITE ..........   98
   10.4.1     UDP .................................................   98
   10.4.2     TCP .................................................   99
   10.5       Reliability for INVITE Requests .....................   99
   10.5.1     UDP .................................................  100
   10.5.2     TCP .................................................  101
   10.6       Reliability for ACK Requests ........................  102
   10.7       ICMP Handling .......................................  104
   11         Behavior of SIP User Agents .........................  104
   11.1       Caller Issues Initial INVITE Request ................  104
   11.2       Callee Issues Response ..............................  104
   11.3       Caller Receives Response to Initial Request .........  105
   11.4       Caller or Callee Generate Subsequent Requests .......  105
   11.5       Receiving Subsequent Requests .......................  106
   12         Behavior of SIP Proxy and Redirect Servers ..........  107
   12.1       Redirect Server .....................................  107
   12.2       User Agent Server ...................................  107
   12.3       Proxy Server ........................................  107
   12.3.1     Proxying Requests ...................................  108
   12.3.2     Proxying Responses ..................................  108
   12.3.3     Stateless Proxy: Proxying Responses .................  108
   12.3.4     Stateful Proxy: Receiving Requests ..................  109
   12.3.5     Stateful Proxy: Receiving ACKs ......................  109
   12.3.6     Stateful Proxy: Receiving Responses .................  109
   12.3.7     Stateless, Non-Forking Proxy ........................  110



Handley/Schulzrinne/Schooler/Rosenberg                      [Page 177]

Internet Draft                    SIP                      July 13, 2000


   12.4       Forking Proxy .......................................  110
   13         Security Considerations .............................  115
   13.1       Confidentiality and Privacy: Encryption .............  115
   13.1.1     End-to-End Encryption ...............................  115
   13.1.2     Privacy of SIP Responses ............................  117
   13.1.3     Encryption by Proxies ...............................  118
   13.1.4     Hop-by-Hop Encryption ...............................  118
   13.1.5     Via field encryption ................................  118
   13.2       Message Integrity and Access Control:
   Authentication .................................................  119
   13.2.1     Trusting responses ..................................  122
   13.3       Callee Privacy ......................................  123
   13.4       Known Security Problems .............................  123
   14         SIP Authentication using HTTP Basic and Digest
   Schemes ........................................................  123
   14.1       Framework ...........................................  123
   14.2       Basic Authentication ................................  124
   14.3       Digest Authentication ...............................  124
   14.4       Proxy-Authentication ................................  125
   15         SIP Security Using PGP ..............................  126
   15.1       PGP Authentication Scheme ...........................  126
   15.1.1     The WWW-Authenticate Response Header ................  126
   15.1.2     The Authorization Request Header ....................  127
   15.2       PGP Encryption Scheme ...............................  129
   15.3       Response-Key Header Field for PGP ...................  130
   16         Examples ............................................  130
   16.1       Registration ........................................  130
   16.2       Invitation to a Multicast Conference ................  132
   16.2.1     Request .............................................  132
   16.2.2     Response ............................................  133
   16.3       Two-party Call ......................................  134
   16.4       Terminating a Call ..................................  136
   16.5       Forking Proxy .......................................  137
   16.6       Redirects ...........................................  141
   16.7       Negotiation .........................................  143
   16.8       OPTIONS Request .....................................  144
   A          Minimal Implementation ..............................  145
   A.1        Transport Protocol Support ..........................  145
   A.2        Client ..............................................  145
   A.3        Server ..............................................  146
   A.4        Header Processing ...................................  146
   B          Usage of the Session Description Protocol (SDP)
   ................................................................  146
   B.1        Configuring Media Streams ...........................  147
   B.2        Setting SDP Values for Unicast ......................  149
   B.3        Multicast Operation .................................  150
   B.4        Delayed Media Streams ...............................  150
   B.5        Adding and Deleting Media Streams ...................  150



Handley/Schulzrinne/Schooler/Rosenberg                      [Page 178]

Internet Draft                    SIP                      July 13, 2000


   B.6        Putting Media Streams on Hold .......................  151
   B.7        Subject and SDP "s=" Line ...........................  151
   B.8        The SDP "o=" Line ...................................  151
   C          Summary of Augmented BNF ............................  152
   C.1        Basic Rules .........................................  154
   D          IANA Considerations .................................  156
   E          Changes from RFC 2543 ...............................  157
   F          Changes Made in Version 00 ..........................  157
   G          Changes To Be Made ..................................  163
   H          Acknowledgments .....................................  168
   I          Authors' Addresses ..................................  168
   J          Bibliography ........................................  169







































Handley/Schulzrinne/Schooler/Rosenberg                      [Page 179]