home *** CD-ROM | disk | FTP | other *** search
- Path: icefog.polarnet.com!not-for-mail
- From: floyd@icefog.polarnet.com (Floyd L. Davidson)
- Newsgroups: comp.dcom.modems
- Subject: Re: 28.8 baud & telco responsibilities
- Date: 25 Mar 1996 19:05:05 -0900
- Organization: _________
- Message-ID: <4j7qdh$f84@icefog.polarnet.com>
- References: <todamhyp-2802961822500001@bhppp31.bluehawk.com> <4it4ol$7ia@castle.nando.net> <4it7e5$b9g@news2.cais.com> <Dou7A8.E4@giskard.demon.co.uk>
- Reply-To: floyd@polarnet.com
- NNTP-Posting-Host: icefog.polarnet.com
-
- In article <Dou7A8.E4@giskard.demon.co.uk>,
- Dale Shuttleworth <dale@giskard.demon.co.uk> wrote:
- >Floyd Davidson (floyd@polarnet.com) wrote:
- >: You can expect a minimum bandwidth of 2400 Hz (from 400Hz to 2800Hz),
- >: and a signal to noise ratio of 24 dB. It won't take too long to
- >: relate that to v.32bis protocols!
- >
- >Ahh, but do they specify the performance with respect to phase? A
- >line which suffers random phase shifting could meet these standards
- >but most QAM (e.g. V34, V32, etc) modems would fail to connect at
- >all. (This is not an uncommon problem!)
-
- There, of course, are specs for phase hits and phase jitter, but that
- misses the point. Both are noise.
-
- >: The maximum bps is directly derived from bandwidth and S/N, and
- >: in practice matches theory.
- >
- >But there is the assumption that those two values completely specify
- >the performance of the line. This is not the case.
-
- With respect to the maximum bps attainable over a voice channel,
- it is indeed totally specified by the bandwidth and the S/N.
- (Keeping in mind that virtually every other parameter will
- affect one or the other of those to the degree that it affects
- the maximum data rate.)
-
- >I expect you could argue that lines where compression is used meet
- >a 2400Hz bandwidth and 24dB SNR but again, the average modem will
- >not be happy.
-
- The average modem _will_ in fact be happy if there really is
- 2400Hz bandwidth and a 24 dB S/N (at least it will be happy
- enough to get 14.4Kbps connections). The biggest problem with
- that thought is most methods for measuring either of those
- parameters are not accurate measurements of the environment that
- the modem operates under. For example the typical methods of
- measurements for a voice channel are very different from the
- methods used on radio systems to measure the identical
- parameters. Noise loading of the entire system is the only way
- to get meaningful measurements, and we don't normally do that
- with VF channels, but commonly do with radio systems. The
- reason is that until recently (the advent of so many high speed
- modems) there was little significance to the difference as far
- as quality goes for a VF channel. But for radio channels (on
- analog radio systems) it has long been known that anything less
- than noise loading is nothing but a hopefully indicative weather
- vane, and not a definitive test.
-
- Floyd
- --
- Floyd L. Davidson floyd@polarnet.com
- Salcha, Alaska or: floyd@ims.alaska.edu
-