[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: Comparisons
>In a study we are finishing we note that there are no industry
>wide standards for ISDN for compression.
True and we need them!
>
>This means, unless you have identical equipment employing the same
>compression scheme at both ends of your ISDN line, you will get
>only raw 128K.
>
>Now take any v.34 28.8 modem, this does have a compression standardd
>and at a compression of 3:1, the through put is 86 kbps. This
>is very close to 128.
>
>So, I suggest for the push for ISDN to be meaningful, someone
>better pay attention to standardizing compression.
While I agree 100% with Curtiss about the need for BRI Terminal Adapter
compression standards, I take issue with his simplistic presentation of
these facts and I'm quite surprised considering this is an MIT study we're
talking about!
Assumption: the application is connection to an ISP.
1. Compression to 86kbps is not at all a given (even statistically) with
v.34. It depends, of course, on the binary makeup of the data; Text
compresses to the max, downloaded .zip files run at about 1:1 - on a _clear_
day. Web graphics also run closer to 1:1 than they do to 3:1.
Granted, PPP efficiencies (or lack of;) muddy this but let's take that as a
common factor.
In fact, given my experiences, I'm not sure there's even much compression
benefit at all - connecting to PPP through v.34.
2. Seems to me that 86kbps is much closer to 64 than to 128; given the
above, I'll take a clear channel 64kbps over 28.8 with compression anyday!
In addition, v.34 only addresses async format. Non ISP applications such as
synchronous gateways (Virtual network links) e.g. LAN - LAN connections
would not benefit; so vanilla bandwidth offered by BRI, along with the very
fast call setup time, still beats the heck out of modem technology.
Correct me if I'm wrong, of course.
Jim Armstrong
- Office Information Technologies, Inc -
513.864.5385
PO Box 2025
Fairborn, OH 45324