bakken@tahoma.ARPA (Dave Bakken) (09/13/87)
I have read of various 9600 baud modems with different (noncompatible) protocols, but have not heard anywhere if any of these is becoming a standard. I even read something on USENET from fidonet about 6 months ago describing their efforts to establish a 9600 baud standard, but haven't ran across anything else about it. So is there a clear winner emerging? I'd hope this question wasn't answered last week as I don't get to monitor this group all the time. We might get a 9600 baud modem in our next budget but need to know the score. Please don't reply if you have a financial interest in any of the involved companies. Thanks! Dave Bakken Boeing Commercial Airplane Company uw-beaver!ssc-vax!shuksan!tahoma!bakken Disclaimer: These are my own views, not my employers.
mw3s+@ANDREW.CMU.EDU (Martin Weiss) (09/15/87)
There are two standards for 9600 bps modems: V.29 for leased line modems and V.32 for dialup. V.29 has been around for a number of years (since 1976, I believe) and V.32 has been around since 1984. V.29 is well entrenched as a 9600 bps leased line standard. It has been adapted by some manufacturers to dialup use by using it in a pseudo-duples mode. V.32 is a very new standard. It uses echo cancellation to separate the originate and receive channels and trellis coding for performance improvement. Full V.32 modems are expensive at this time, because the chip sets for echo cancellation aren't cheap. Some manufacturers use a sub-set of V.32 for their modems, and others use proprietary techniques. Hayes, for example, does not adhere to V.32 to the best of my understanding. My guess is that V.32 will emerge as the dominant standard in dialup 9600 bps modems, but that it will take a while. This will be due in large part to the cost of implementing echo cancellation.
taylor@ecsvax.UUCP (Steven Taylor) (09/16/87)
There is an additional problem with any 9600 bps dial-up modem. Many organizations are now converting their internal voice transmission systems to T-1 based systems using compressed voice. Unfortunately, virtually none of the compressed voice algorithms support 9600 bps modem transmission. The systems are fine for voice, but modems, especially those at high speeds, use the available spectrum very differently than "normal" voice. A handful of manufacturers have some specialized techniques to allow V.29 transmission with digital voice at 32 kbps (standard telco uses 64 kbps PCM), but these are proprietary to each vendor. I haven't heard of any which support V.32. This is a MAJOR problem for systems designers. With systems offering up to 10 to 1 compression over the telco standards, it's hard to avoid using compression of voice. However, it does cause this problem for "normal" modem use. As we approach an "all digital world," this could be a major stumbling block within many organizations, so, personally, I am not too sure what the future of the high speed modems will be. I guess it depends a lot on the extent to which the digital services are extended, and at what speeds. For a related problem with V.32 and use on "OCC's" (not AT&T), see the news item in the September "Data Communications" magazine, page 15. (Is this the appropriate newsgroup for ISDN discussions? I've been off the net for a while, so I'm not sure. Thanks.) Steve Taylor Distributed Networking Associates Greensboro, NC 919-292-4444