Strawman use cases for matching rule discussion

Matching rules can operate in complex and sometimes surprising ways, so it helps to have a set of use cases to analyze, to make it easier to understand what a given matching rule proposal will produce. 

Even though there are outstanding issues relating to how FEC/RTX/RED is handled in ORTC, I believe that it still makes sense to include this in the use cases, so as to avoid unpleasant surprises down the road (e.g. oops... FEC/RTX/RED doesn't work).

Here are a few potential use cases that come to mind.  Comments/additions welcome. 

Audio use cases:   
1. RtpReceiver is set up to support more than one audio codec with a distinct payload type specified for each codec. 
   The sender SSRC is not provided during signaling. Goal is to have a single audio track as output. 

2.  RtpReceiver is set up with a single audio codec with a pre-provisioned payload type.  The sender SSRC is not provided 
    during signaling.  Due to an SSRC conflict, the SSRC changes.  Goal is to switch seamlessly between SSRCs with no unhandledrtp event. 

3. RtpReceiver is set up with a single audio codec for an audio conference with the bridge forwarding traffic from up to 3 speakers 
   (based on audio levels).  A single PT is used for all streams.  The bridge may either:
      a. allocate its own SSRCs which can be provided during signaling or 
      b. forward audio packets (RTP translator) so that the SSRCs are not provided during signaling. 

In either case a) or b), the receiver may assign muxIds to each of the 3 speakers which it expects the sender to use. 
 
Video use cases: 

4. RtpReceiver is set up to receive simulcast from an SFU at one of 2 potential resolutions, each using the same payload type.
   There are also FEC streams associated with each simulcast streams, operating on a pre-provisioned payload type.  
   The Selective Forwarding Unit (SFU) may either: 

   a. Allocate a distinct SSRC for each simulcast and FEC stream or
   b. Use a single SSRC for both simulcast streams, as well a single (distinct) SSRC or both FEC streams. 

5. RtpReceiver is set up to receive SVC utilizing temporal scalabiltiy from an SFU.  Each temporal stream uses the same payload type.
   There is also an RTX stream on a different payload type which protects the base layer only.  The SFU may either: 

   a. Utilize the same SSRC for each of the layers or
   b. Utilize a distinct SSRC for each of the layers.  

 

Received on Friday, 5 December 2014 23:04:19 UTC