Relay-Version: version B 2.10 5/3/83; site utzoo.UUCP Path: utzoo!mnetor!seismo!decuac!hadron!cos!howard From: howard@cos.COM (Howard Berkowitz) Newsgroups: comp.dcom.modems Subject: Re: Multiplexor Advice Message-ID: <180@cos.COM> Date: Thu, 19-Mar-87 07:56:38 EST Article-I.D.: cos.180 Posted: Thu Mar 19 07:56:38 1987 Date-Received: Sat, 21-Mar-87 03:13:01 EST References: <116@medivax.UUCP> <1901@hoptoad.uucp> Distribution: na Organization: Corporation for Open Systems, McLean, VA Lines: 35 Keywords: HELP Summary: Two-level multiplexing usually is best. > > I am also interested in any information concering multiplexors. Our > radio station is tied with our sister station in San Jose through a T1 carrier. > I have one extra channel that I wish to plug a multiplexor into. The unit is > built by Rockwell (you know, the shuttle Chalenger people). Its compatible > with units built by Coastcom. I'm interested in multiplexing 6 9600b/s paths > (6*9600=56000). Any suggestions out there? Has any one had any problems I > should look for? I'm unclear if your extra channel is another T1 or a "56" KBPS channel. (The 1.544 MBPS rate, technically, is a DS1 signal, while its subrate 64 KBPS channels, often bit-stuffed to 56KBPS transparency, are DS0. "T1", technically, is the Bell System (or nonsystem) code for DS1 signals over repeatered twisted pair, as opposed to other transmission media such as microwave or fiber). If your channel is DS0, you're going in the right direction; it's much cheaper to use 2-stage multiplexing to get a relatively slow speed (in DS1 terms) such as 9600 BPS. You rarely want an intelligent (statistical) multiplexer on DS1; your approach to engineering the system should assume transmission speed is essentially free. A fixed-slot bit-interleaved multiplexer usually is best (see GDC, Timeplex, etc.). If you are not running synchronous protocols on the 9600, my recommendations are likely to change. Let me know what you are running and I'll try to help. When buying the mux, be sure your mux vendor works with you and the T1 mux people to understand where all the clocks are coming from, and how they stay locked to each other. This can be nontrivial. Howard Berkowitz (703) 848-2100.