Re: [AVT] How is SDP a=fmtp used for RFC 2198?

Hi Gunnar,

I thought of one more thing in relation to congestion control. I think 
the below is okay for real-time typing. However one should most likely 
make further comments that if one uses this payload format to transport 
larger blocks of text (automated systems), there is further 
considerations to be made in relation to using redundancy. Then it might 
be better to use either adaptive FEC or redundancy to combat packet loss 
due to the then higher overhead for each redundancy level. The main 
point is that the is important to perform congestion control, and the 
actual impact needs to be considered depending on application.

cheers

Magnus

Gunnar Hellstrom wrote:

> Fine,
> 
> I think the advice to maintain redundancy level, but increase the
> interval
> between transmissions of text in a congested situation is good.
> 
> 1. We have 300 ms between transmissions as the default. ( that makes it
> quite burst-safe and keeps total load from text at a low level. )
> 
> 2. We also agreed to having one original and two redundant transmissions
> as
> default.
> 
> 
> T.140 requires the wait for transmission to be less than 500 ms. So,
> there
> is a room to extend the interval between transmissions in congested
> situations to 500 ms even without breaking the standard. The users will
> experience a slightly more jerky but still acceptable text presentation.
> 
> We will introduce this reasoning in next revision of RFC2793bis to be
> issued soon.
> 
> Gunnar
-- 

Magnus Westerlund

Multimedia Technologies, Ericsson Research EAB/TVA/A
----------------------------------------------------------------------
Ericsson AB                | Phone +46 8 4048287
Torshamsgatan 23           | Fax   +46 8 7575550
S-164 80 Stockholm, Sweden | mailto: magnus.westerlund@ericsson.com

_______________________________________________
Audio/Video Transport Working Group
avt@ietf.org
https://www1.ietf.org/mailman/listinfo/avt