W3C home > Mailing lists > Public > public-webrtc@w3.org > July 2011

Re: Symmetrical streams

From: Magnus Westerlund <magnus.westerlund@ericsson.com>
Date: Mon, 11 Jul 2011 22:18:05 +0200
Message-ID: <4E1B5A7D.9000501@ericsson.com>
To: public-webrtc@w3.org
Hi,

I think Harald's comment needs a bit of clarification.


On 2011-07-11 10:12, Harald Alvestrand wrote:
> On 07/01/11 05:12, jywang wrote:
>> Hi stefan,
>> 
>> I think a PeerConnection object maybe is good to group streams
>> which should be transmited in the same 5-tuple, and another object
>> can be created if need.
>> 
>> In this way, the developer can do their selection, and the service
>> provider also understand how to decide. However, we have to
>> constrain the PeerConnection's concept for this purpose.
> I don't think the PeerConnection today is a good description of
> mapping to RTP sessions.
> 
> The currently accepted wisdom in IETF is that between two peers,
> there should be one RTP session containing all the video media, and
> one RTP session containing all the audio data. Whether the two RTP
> sessions are distinguished by 5-tuple or by some other means, and if
> there really needs to be two, is still a debate.

The one audio, one video RTP session is correct for basic use cases.
However, for some more advanced use cases, for example within video
conferencing can result in that an RTP end-point has one or more RTP
session per media type. Each RTP session can contain zero to many media
sources, e.g. encodings from different video cameras, where each source
is an SSRC. A particular end-point may therefore both receive multiple
SSRCs (media streams) and send multiple media streams.

> 
> It's logical for these RTP sessions to be bidirectional.

RTCP within an given RTP session should always be bi-directional for
efficiency in number of transport flows needing to be established and
NAT traversal. Some application usages will result in that some RTP
session will only be unidirectional when it comes to actual media
streams. And uni-directional media streams of different purposes should
not be group together into a common RTP session over the same
bi-directional transport just to create bi-directional flows as that
messes with RTCP behavior.

Cheers

Magnus Westerlund
Chair of AVTCORE and AVTEXT WGs in IETF, the maintainer and extenders of
RTP.


----------------------------------------------------------------------
Multimedia Technologies, Ericsson Research EAB/TVM
----------------------------------------------------------------------
Ericsson AB                | Phone  +46 10 7148287
Färögatan 6                | Mobile +46 73 0949079
SE-164 80 Stockholm, Sweden| mailto: magnus.westerlund@ericsson.com
----------------------------------------------------------------------
Received on Tuesday, 12 July 2011 10:00:16 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Tuesday, 12 July 2011 10:00:25 GMT