Hi Dale,

Thanks for the response..
I know the only thing that we can do for this scenario is to have B2BUA 
sends REFER method so that user A can initiate new INVITE and hence it can 
act as UAC and respond to 180 ringback.

Even if we use INVITE with Replaces, it will still act as UAS and hence 
will not process 180 response.

Actually only concern is because of the exsiting architectute B2BUA has to 
keep track of different dialogs which is not generally the case in other 
protocols if he wants to implement these types of scenarios. Some 
implementations think it of as call legs (or sessions) and they dont want 
maintain two different dialogs which are more or less session for them.

As there is no mechanism to reinitialize the state of the dialog of UAC, 
you always need to send the REFER method so that UAC start a new dialog 
and hence will behave as a new call. It is a new dialog but for user its 
not a new call.

One possible quick solution was to have Aler-Info header in the Re-invite 
send to the UAC which will indicate to UAC that UAC can play ringback to 
the user, but this only solves the ringback problem. There can be other 
different scenarios.

Actually the scenario which I have discussed is very common, it can be for 
toll free calling card calls also where first the call is connected to 
voice mail server (using 2xx and not reliable provisional response) to get 
the digits and then call is connected to the actual destination party.
Basically, from SIP perspective we have to treat all these scenarios as 
like transfer where B2BUA takes the role of transferor and connecting the 
user A to different calls. I am not sure if this type of approach will 
have any repurcussions on billing issues.

-Udit




"Dale Worley" <[EMAIL PROTECTED]> 
Sent by: [EMAIL PROTECTED]
04/21/2005 12:30 PM

To
<[email protected]>, <[email protected]>
cc

Subject
[Sip] RE: [Sip-implementors] UAC behaviour






> From: [EMAIL PROTECTED]
>
> Consider the following prepaid card application scenario::
>
> Assuming party A and B are talking and following steps occur:
> i)   Party B hung up.
> ii)  B2BUA end the dialog with user B and connect user A to
> media server
> using REINVITE asking for another number.
> iii) After A presses the digits, B2BUA initiates a new dialog
> for user C
> iv) When party C is ringing,  B2BUA sends 180 Ringing to agent A.

IMHO, the fundamental problem is that your call flow maintains one dialog
between A and the B2BUA, but you want to reflect back to A all the call
progress indications (provisional responses) from the dialogs between 
B2BUA
and B, C, etc.

To avoid this, perhaps the B2BUA could issue INVITE with Replaces: to A to
"reinitialize" the dialog state with A.  (In reality, replace the previous
dialog with a new one.)

>
>
>     User A              B2BUA               User B
> Media Server
>       |                   |(1) BYE            |                   |
>       |                   |<------------------|                   |
>       |                   |(2) 200            |                   |
>       |                   |------------------>|                   |

|  INV with Replaces|
|<------------------|
|  BYE of dialog 1  |
|------------------>|
|  200 of dialog 2  |
|------------------>|

OK, I'm too lazy to draw the entire message flow.  But it seems like you
could generate suitable INVITEs with Replaces and/or REFER messages to 
cause
A to drop the dialog with B, establish a new one with the media server, 
then
drop that dialog and replace it with one to C, etc.  Since each is a new
dialog, A's UA will receive and display correctly any call progress
indications.

This is also more in line with the architecture envisioned in RFC 3261.
Instead of having a B2BUA, you might be able to just use a proxy that 
keeps
itself in the signaling path with a Record-Route.

Dale


_______________________________________________
Sip mailing list  https://www1.ietf.org/mailman/listinfo/sip
This list is for NEW development of the core SIP Protocol
Use [email protected] for questions on current sip
Use [email protected] for new developments on the application of sip
_______________________________________________
Sip-implementors mailing list
[email protected]
http://lists.cs.columbia.edu/mailman/listinfo/sip-implementors

Reply via email to