56 kbps modems

[Jay Gairson]

I personally, am not much for this, I am just one with various ideas, and 
questiosn and such, so here goes.

On Mon, 16 Sep 1996, Remo Pini wrote:

> -----BEGIN PGP SIGNED MESSAGE-----
> 
> Mime-Version: 1.0
> Content-Type: text/plain
> Content-Transfer-Encoding: 7bit
> 
> To: cypherpunks at toad.com
> Date: Mon Sep 16 15:59:53 1996
> I think I have to throw in some words here, too:
> 
> 1. BAUD is one distingushable signal form.
> 2. BPS is bits per second

Ok, I agree with those.

> 
> 3. NO transmission over a standard phoneline can have more than 3100 baud, 
> because the frequency of anything transmitted over that line is band 
> limited to 300hz - 3400hz. (If you have ISDN it's not relevant anyway, 
> since you are fixed with 8000hz or 64000 bps->in europe :), 56000 bps in 
> usa)

Personally, by saying that no transmission on a standard phoneline can 
have more than 3100 baud, is a statement saying that technology and 
science, shall never advance, to a point where things are possible.
If you remember right, just 10 or so years ago, we stated that 2400 baud 
was the highest possible baud, and we would never go over.  My modem, 
running on a standard phone line, is a 28.8 USRobotics modem, with 
the software, and hardware upgrades to a 3400 baud, I get on an average 
day anywhere from 3500 to 3600 baud for send/receive.  On a bad day, I 
only get 3100 to 3200.  SO I would say, that my phone lines, are cleaner 
than most eh?  And Yes, I am in the US.
Speaking of ISDN, how many people, can afford to have a personal ISDN 
line in there house?  And then afford to connect to something/someone 
else on a next to permanent basis monthly?

> 
> 4. Most modern transmission schemes work with multiple bits per baud. I.e. 
> you transmit 10 bits in one baud if you have a 31000 bps modem. the only 
> limitation in transmission speed is the amount of binary values you can 
> pack into one baud. that on the other hand is limited by the S/N (signal to 
> noise) ratio of your line. If you have a noise of 0.9%, you can't use more 
> than 100 steps or you have ambiguous signals. since people talk about 56000 
> bps modems (we tried 34000 modems here and they couldn't produce more than 
> 28800 on a very good connection) that would mean, that you have to transmit 
> 18 bits = 262144 (!) distinguishable signal forms per baud.
> 
> Comments?

What about new ways of splitting the steps and baud more so that it shows 
less at a higher level....  Just a question..  Ahh well, I'm getting a 
page so I shall finish this now...  Answer appreciated..

> 
> Remo Pini
>  
> - ------< fate favors the prepared mind >------
> Remo Pini                        rp at rpini.com
> PGP:  http://www.rpini.com/crypto/crypto.html
> - ----< words are what reality is made of >----
> -----BEGIN PGP SIGNATURE-----
> Version: 2.6.3i
> Charset: noconv
> 
> iQEVAwUBMj1dWhFhy5sz+bTpAQHmVAgAqRHydyZdElXlwMSgdKB6hg1fDBCMfhia
> JoI5W8n0JLcKe9HYN/H0DPqFHHDIATEhlN3b3OIhYCw52cNJ/e3b9Nbp5RQo+sDX
> zAogz5wZiDV7EA/gL589lNQZ9VKHlgTYBLzu1tqyJ5cD2KhWEUjvXyN4lYuxcEQT
> NIaiNPeYXrC0BeoLa/AE8mCrtu+7nhxy5HlSjDiu3lEYaVygKIPQHM3+Ljzq0jkq
> bmbqJbTyZshos+5gxHyXLsbL8rkpST53YT4Z3clL6PCv1ntNGXtb/loWgIDallYJ
> bHgfh/bAS5Utg7fpzuVNS8AJV8L2VLirScBd4Bq5RAXZxnoTVVQxOw==
> =9JSI
> -----END PGP SIGNATURE-----
> 



Erp