Modem development question...

  • Thread starter Thread starter rdlebreton
  • Start date Start date
R

rdlebreton

Okay,

This may seem like a lame question but...

Is ANYONE doing any research to increase the performance of a
"dial-up" modem beyond the so-called 56kbps limit?


TIA
 
rdlebreton said:
Okay,

This may seem like a lame question but...

Is ANYONE doing any research to increase the performance of a
"dial-up" modem beyond the so-called 56kbps limit?
Click to expand...

With the speading use of broadband it wouldn't be considered a good
investment
 
rdlebreton said:
Okay,

This may seem like a lame question but...

Is ANYONE doing any research to increase the performance of a
"dial-up" modem beyond the so-called 56kbps limit?
Click to expand...

If I recall correctly, speeds above 28.8 (or maybe even 14.4) are not a
consequence of more rapid transmission, but from compression of the
data. Because the compressibility of the data is dynamic, I suspect
that means that 56K is possible, but not sustainable in practical use.
Perhaps better compression schemes will come along, but I wouldn't hope
for a dramatic improvement.

Another issue is line quality. I can't really say much about the
practical limit imposed by line quality, other than there is one.
 
rdlebreton said:
Okay,

This may seem like a lame question but...

Is ANYONE doing any research to increase the performance of a
"dial-up" modem beyond the so-called 56kbps limit?
Click to expand...
the real limitation is the line itself...
 
Your question suggest you don't yet know the basic science
nor the reason why these limits exist. In simplest times, the
naive complained that copper wire was a speed limiting
factor. They were wrong. Speed limiting factor for modems
(on POTS systems) is the switching computer inside the telco's
toll station. The fundamental science was defined by Claude
Shannon in the Bell Labs in 1948 (back when the Labs were run
by people who had science backgrounds). To you, this means
that 56K will be the best speed theoretically possible without
completely changing hardware in that switching station
computer.

BTW, what made 56K modems able to do up to 53K? Eliminate
only some hardware in the switching computer - an A/D
converter.

Of course Clayton Christensen defined the underlying concept
in his book Innovator's Dilemma. Disruptive innovations
simply change your question to be myopic and irrelevant.
British Telephone was demonstrating DSL in 1981. IOW long
before even 56K existed the telephone switching computer was
being obsoleted by a disruptive technology. That means
scrapping the entire $multi-million computer that most
switching stations only replaced in the 1980s and 1990s. You
can see why Baby Bells so feared broadband and why Congress
had to pass the 1996 Communications Act to force them to
innovate.

Isenberg best defined this in his AT&T Bell Labs paper about
the 'Smart and Dumb' networks. The Dumb network being a
superior solution. But again, AT&T so routinely stifled
innovation that Isenberg could not even put that paper on his
own web site. You can find it through http://www.isen.com .
IOW that 'easy to read' paper may better demonstrate why your
56K modem question is about promoting obsolete technology.

I can't say enough about the concepts promoted by both
Isenberg and Christensen. Concepts that too many computer
users still don't appreciate. But underlying your original
question are technical limits defined by Shannon's epic paper
"A Mathematical Theory of Communication". So understated and
yet so revolutionary to digital communication.
 
I recall that for a long time 1200 baud modems were considered the
maximum speed theoretically possible, then 2.4, 14.4, 28.8 and then 56k.
It is somewhat off-point to talk about central-office equipment since
dialup modems are POTS devices; only the ISP's CO gear needed upgrading
to support 56K, not the end users'. Did Shannon actually say that 56k
was the best POTS could do? Speeds have gone up as compression and
modulations schemes have been improved (and the cost of implementing
them has gone down.)

I agree that there's not much incentive to up POTS modem speed with DSL
now available though I'd guess that lots of work is always being done on
better modulation and compression schemes for all sorts of wired,
wireless and optical communications.
 
rdlebreton said:
This may seem like a lame question but...

Is ANYONE doing any research to increase the performance of a
"dial-up" modem beyond the so-called 56kbps limit?
Click to expand...

It's only lame if you don't know the underlying facts. Most
telephone service today is conducted over digital links, operating
at 56kb. There is no added capacity available. The modem is
already doing magical things by getting your signal over the highly
variable POTS pair to the telephone exchange, where that voice
signal is converted to digital. The modem has been able to figure
out what to do to get that capacity over that 2 wire analog pair.
Going any further won't do anything.

That is why you don't get 56 kb connections if you live too far
from the exchange (or the exchange isn't digital), and also why the
actual limit is about 53 kb (because otherwise excessive power
would be needed for the transmission, and exceed FCC regulations).
BTW, the specified bandwidth on that POTS line is about 3.5 kHz.

So the answer is that, given the existing telephone network, higher
speeds are impossible.

--
Some useful references about C:
<http://www.ungerhu.com/jxh/clc.welcome.txt>
<http://www.eskimo.com/~scs/C-faq/top.html>
<http://benpfaff.org/writings/clc/off-topic.html>
<http://anubis.dkuug.dk/jtc1/sc22/wg14/www/docs/n869/> (C99)
<http://www.dinkumware.com/refxc.html> (C-library}
<http://gcc.gnu.org/onlinedocs/> (GNU docs)
 
This is where wikipedia is better first consulted.
http://en.wikipedia.org/wiki/Information_theory

Signal to noise ratio is one concept. To make 56K possible,
the noise from one A/D converter was eliminated. So the sever
had to make a direct connection to a telco switching station
without that A/D converter. IOW hardware had to be removed
from the switching computer to make 56K possible - to
eliminate a source of noise - to meet criteria defined by
Shannon.
http://en.wikipedia.org/wiki/Analog-to-digital_converter
 
Back
Top