home *** CD-ROM | disk | FTP | other *** search
- Path: ctsasync62.cc.utah.edu!marc.fuller
- From: marc.fuller@m.cc.utah.edu (Marc Fuller)
- Newsgroups: comp.dcom.modems
- Subject: Re: 512Kbps modem developed by Ericsson
- Date: Thu, 18 Apr 96 05:06:01 GMT
- Organization: U of U
- Message-ID: <marc.fuller.1180191601A@news.cc.utah.edu>
- References: <4k9grj$q2t@rubens.telebyte.nl> <4k9qem$fnf@sam.inforamp.net> <pjkDpw55L.Czt@netcom.com> <3172f9fb.2406869@news.pbinet.com> <pjkDpyArv.sF@netcom.com>
- NNTP-Posting-Host: ctsasync62.cc.utah.edu
- X-Newsreader: VersaTerm Link v1.1.6
-
- >
- >
- >I admit that with _current technology_ it would be hard to imagine getting
- >much more speed out of the existing infrastructure. That being said,
- >head in the sand science doesn't go anywhere either. I'm sure you are aware
- >of the kinds of things DSP has done that "traditional wisdom" claimed couldn't
- >be done a few years ago.. i.e. current levels of areal density achieved in
- >modern disk drives using PRML read channels, etc. Nyquist and Shannon have
- >their theorems, but that's just what they are: theories that are true until
- >someone either proves them wrong or comes up with a creative way to circumvent
- >them. I'm sure you're aware of superconductivity, another "impossible"
- >not that long ago. And its well-known in audiophile circles that the
- >human ear can perceive sound beneath the supposed "noise floor" that the
- >quasi-scientists would have everyone believe was the "brick wall" beyond
- >which no information could be extracted. And so it goes.
- >
- As someone with a background in statistical communication theory I find what
- you
- are saying is not too believable. Most of the things that you mention were
- proven
- wrong because we did not understand the physical properties of either energy or
- matter. Shannon's theorem is based purely on probability. It says that for a
- given signal to noise ratio in any medium there is a fundamental limit on
- how much information can be transmitted. As someone that actually tried to
- increase the information flow on the customer loop there is only so much you
- can do before you bump up against those limits. My guess is that this
- product either raises the bandwidth or power on the customer loops.
- Unfortunately in the real world their are strict limits on this because of
- crosstalk to other customers and power limitations. What you are proposing
- is a violation of basic mathematical principals. Unless you are planning to
- rewrite those I have a hard time believing that their will be a great
- advance in the limit of information transmission on the customer loop. The
- difference between now and previous develpments in modem technology is that
- before we were up against technological limits (couldn't figure an easy or
- cheap way to do something). Now we are close to the theoretical limits.
-