A collegue and I are having a bit of a debate as to which compression technique to use on our 64K WAN links. A typical remote site consists of up to two IP Phones and two workstations. The workstations are used for e-mail and http traffic to pull info off of our servers back at the main office. The 64K links connect to a few distribution routers around the continent and then each use a T-1 back to the main site. The voice is okay, but I would like to make it a bit better by using rtp header-compression. He suggested using compress stac and just compress the whole ball of wax. Are there any pros or cons of each?
Here is another twist, a couple other links are 48K sattelite links. The voice was horrible until I added the header-compression, but the same compression debate exists with those.
One more thought, fast-switching is supported with rtp header-compression, but not with stac. It looks to me like it has become one of those "lesser of the two evils" type scenarios and would like some additional opinions from the experts or anyone else who has experimented with something similar.
All WAN links use HDLC and that is not an option to change at this time.
Any thoughts, comments, or opinions are appreciated.
Thanks,
Mike
