Re: [Sip-implementors] Definition of Call-ID

[Hemant Pandey]

Hi All,       These are the following points which i want to discuss.   1. 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.  If the request is INVITE and the final response is a non-2xx, the transaction also includes an ACK to the response.  The ACK for a 2xx response to an INVITE request is a separate transaction."   As per this definition of SIP Transaction in RFC3261(page No.-24), i could understand that SIP Transaction has peer-to-peer significance.It means SIP Transaction is established, when UAC receives final response (2xx or some error) from final UAS. Contradictory to this, while sending INVITE message from UAC to final UAS, each stateful proxy  in the route adds Via header field, which has a branch parameter.This branch parameter identifies the transaction. Then why these branch parameter has different values? It means in a call SIP Transaction is established corresponding to each SIP element in the route. Please clear this doubt.   2. As i understood, one SIP transaction Consist of all messages between a Request and its Final Response. In case of REGISTRATION process, UAC sends REGISTER method with particular value of Branch parameter in the Via Header field, and 401 Unauthorized Final response. UAC again sends REGISTER method with same branch parameter, it means same transaction is continued. Similarly the same Call-ID is continued for the whole REGISTRATION process. It means in this case Call-ID and Branch parameter identifies same things. Becase in REGISTRATION process, call-ID is nothing but identifier to group of messages not the identifier of the call. so please let me know, whether i have understood wrong or right?   Thanks for ur cooperation.   Regards    Hemant   ----- Original Message ----- From: Chris Boulton To: Hemant Pandey ; [EMAIL PROTECTED] Sent: Friday, February 28, 2003 3:30 PM Subject: RE: [Sip-implementors] Definition of Call-ID <<< See comments >>>>   -----Original Message----- From: Hemant Pandey [mailto:[EMAIL PROTECTED] Sent: 28 February 2003 03:31 To: [EMAIL PROTECTED] Subject: [Sip-implementors] Definition of Call-ID   Dear All,        To implement multimedia services in 3G network, I recently started studying RFC3261 for SIP protocol. Since i am new to SIP protocol, i have confusion about few terms used in RFC3261. 1. Call-ID   What exactly this Call-ID identifies?   <<<< The Call-ID is seen as a unique identifier for a particular call.  It is then used by the originating client to match responses appropriately.      In its definition in RFC3261, Call-ID acts as a unique identifier to group together a series of messages. So what does this group means?   <<<<A Sip call consists of a series of requests/responses.  Clients (both client and server) need a mechanism to associate(group) such requests/responses.>>>>   2. What is target refresh requests? How these requests could be differentiated with other requests?               <<<< It is a ‘keep alive’ mechanism for sip.  They are differentiated by incrementing the sequence number of the new request (CSEQ).   3. which RFC should be referred for latest SIP call flows examples?               <<<< See draft-ietf-sipping-basic-call-flows-01.txt from the ietf sipping group. >>>>   i will be obliged if any one help to clear these points.   Regards   Hemant Pandey