The MPLS WG Archive

Cell Relay Retreat>MPLS WG Archive>month:2003-Feb> msg00056



[Date Prev][Date Next][Thread Prev][Thread Next]  
  [Date Index][Thread Index][Author Index][Subject Index]

E2E VoIP over MPLS ('VoMPLS') Header Compression

  • From: Curtis Villamizar <curtis@fictitious.org>
  • Date: Wed, 19 Feb 2003 16:20:44 -0500
  • cc: curtis@fictitious.org, "Ash, Gerald R \(Jerry\), ALABS" <gash@ems.att.com>, "raymond zhang" <zhangr@info.net>, "MPLS@UU.net" <MPLS@UU.NET>, "George Swallow" <swallow@cisco.com>, "Loa Andersson" <loa.andersson@utfors.se>, "GOODE, B \(Bur\), ALABS" <bgoode@att.com>, "Dave Cooper" <cooper@GBLX.net>


In message <001001c2d850$57c1d7e0$3a785b87@mt.att.com>, "Jim Hand" writes:
> 	A 50ms queuing point with a G.711 codec will yield packets over 300 byt
> es.
> A 50ms queuing point with G.729x codecs will yield packets under 50 bytes.
> With G.729x with no connection muxing 200-300 byte payload sizes yield
> 200-300ms delays.
> 
> 	This is why I had been assuming, apparently along with some others,
> that you were talking about connection muxing.
> 
> Jim Hand


I wasn't before but you are correct.  200-300ms delays would be
intolerable and you would have to resort to muxing for reasonable
efficiencies.  40 msec queueing point and RTP/UDP header compression
only might be a reasonable interim.

Curtis