I'd like to add my support for the consideration of a modified
checksum to support voice/video applications - it would be interesting
to know what the potential benefits and architecture considerations.
I've worked on various link level protocols - some of which could
easily have supported this - we have a synchronous HDLC driver
we developed back in '94 which did this, based on the 85230 SCC.
The packet modem I'm working on could easily support this, integrated
with the FEC, should it prove useful...
By way of history, I promised to send the reference for the previous
work by CCITT (now ITU-T) who developed an HDLC-based "standard":
The G.754 Recommendation for voice packetisation. (Study group XV).
I believe this was primarily intended for frame relay
using vocoded voice, and specified an HDLC frame format (UIH) which
provided alternate checksumming including the HDLC header and
a ceratin number of pre-agreed bytes of the voice payload header. Alas,
I can't find a copy of the recommendation in my files.
Gorry Fairhurst.
This archive was generated by hypermail 2b29 : Mon Feb 14 2000 - 16:14:56 EST