Re: [rtcweb] An input for discussing congestion control (Fwd: New Version Notification for draft-alvestrand-rtcweb-congestion-00.txt)

[Randell Jesup <[e-mail]>]

On 9/19/2011 4:29 AM, Henrik Lundin wrote:
> Tim,
>
> The problem that we are trying to attack with the proposal is perhaps 
> better described as "available bandwidth estimation" rather than 
> congestion control. I _would_ like to try and separated it from 
> codecs. The bandwidth estimator should inform the sender of the 
> current feasible rate between two endpoints, and it should then be up 
> to the sending client (application, codecs, channel coding, etc) to 
> figure out a way to stay within that limit, while maintaining an 
> acceptable quality for the use-case at hand.

Right - I think we're in agreement (and Tim is) that this is the best 
tack to take on the problem - involve the application in figuring out 
how to meet the bandwidth limits.  This part of it will largely be an 
API specification problem; figuring out how to set up the API to tell 
the JS app what is available, and the interfaces for responding to that.

>
> On Mon, Sep 19, 2011 at 10:21 AM, Tim Panton <[e-mail] 
> <mailto:[e-mail]>> wrote:
>
>
>     I'd like to lurk  on any BoF list, if I may,
>     but I have a feeling that congestion control isn't quite such a
>     self-contained problem
>     as forming a BoF list might imply.
>
>     Roughly, my argument is that any modern congestion control for RTC
>     needs to involve the application
>     and the codec.
>

Absolutely agree.

>
>     Here's an example:
>            I'm in a video conference with 4 other people and I'm
>     exceeding the available downlink
>     bandwidth. What I want is that the video degrades (framerate of
>     quality) without _any_ degradation
>     of the video.
>

I'm a little confused by that wording, but I suspect we agree.  The app 
might want ask the senders to cut their framerate or cut their 
resolution, or cut quality but keep framerate & resolution, or ask the 
senders to switch to lower-bandwidth parameters unless the VAD decides 
the person is talking, etc.  No one answer is the 'right' one for every 
app or situation.

>     I might even want to drop all the video but keep the audio and
>     screenshare alive.
>

Exactly.

>            Conversely if I'm watching an ice hockey game - I'd rather
>     lose audio quality than frame-rate.
>
>     But in _both_ cases I'd like the app to try and tweak the encoder
>     params to get within the available bandwidth
>     before we start injecting delay or dropping frames.
>
>     So my proposal is that the web app is given a chance to get within
>     the available bandwidth
>     before the system starts putting hard constraints on the RTP level.
>
>     All of which makes it something we need to discuss in a holistic
>     fashion with the rest of the API.
>

-- 
Randell Jesup
[e-mail]