[comp.dcom.modems] Why are dial-up modems improving faster than direct-wire ones?

roy@phri.UUCP (Roy Smith) (09/01/89)

	Right now I'm talking to you via a Trailblazer which can do
19.2kbps over plain old dialup lines with something like 3 kHz bandwidth
(and, in fact, can manage a respectable fraction of that over horrible
dialup lines).  On the other hand, I've got a 4-wire leased line (LADC)
circuit at work which I can barely get 128 kbps out of.  It's about a mile
of 26 gauge wire with a measured loop resistance of something like 700 ohms
and useable bandwidth from DC to over 100 kHz.  30 times the analog
bandwidth of a dialup connection, yet I can only coax about 6 times the bit
rate through it.  Not to mention the additional factor of two improvement I
should be able to get because I've got 4 wires.  I've seen a modem which
claims to be able to do 230 kbps over our line, but it doesn't work.  On
the other hand, I've heard of V.32 modems which can do 38.4 kbps over
dialup lines.

	So, my question is why do modern dialup modems get so much better
performance (compared to the channel they have to work with) than leased
line modems?
-- 
Roy Smith, Public Health Research Institute
455 First Avenue, New York, NY 10016
{att,philabs,cmcl2,rutgers,hombre}!phri!roy -or- roy@alanine.phri.nyu.edu
"The connector is the network"

chris@mimsy.UUCP (Chris Torek) (09/01/89)

In article <3971@phri.UUCP> roy@phri.UUCP (Roy Smith) writes:
>... 30 times the analog bandwidth of a dialup connection, yet I can
>only coax about 6 times the bit rate through it. ... I've seen a modem
>which claims to be able to do 230 kbps over our line, but it doesn't
>work.  On the other hand, I've heard of V.32 modems which can do 38.4
>kbps over dialup lines.

They do not work (that way) either.  V.32 modems get 9600 bps in each
direction, no more.  (Sometimes less, when they fall back to 4800 bps,
etc.)  Modems that claim higher rates really shove fancy compression
protocols on the front.  ASCII text can often be compressed to 30% of
its original size, so the apparent bit rate can become 9600/.3=32000
bps (not 38400 bps: that speed is merely required by the doubling rule:
either you run at 19200 bps or you run at 38400 bps, so if you want to
get 32000 bps, you had better run at 38400 bps.)  Trivial text can be
compressed even more, but trivial text does not occur often.

Anyway:

>	So, my question is why do modern dialup modems get so much better
>performance (compared to the channel they have to work with) than leased
>line modems?

Mainly marketing.  It is easier to do fancy compression and DSP at low
rates like 38400 b/s than at 100000 b/s, but the biggest reason is that
there are more telephones out there than leased lines.
-- 
In-Real-Life: Chris Torek, Univ of MD Comp Sci Dept (+1 301 454 7163)
Domain:	chris@mimsy.umd.edu	Path:	uunet!mimsy!chris

henry@utzoo.uucp (Henry Spencer) (09/07/89)

In article <3971@phri.UUCP> roy@phri.UUCP (Roy Smith) writes:
>	So, my question is why do modern dialup modems get so much better
>performance (compared to the channel they have to work with) than leased
>line modems?

Same reason you can buy a zillion decent AT clones but not a single Sun-3
clone:  bigger market.
-- 
V7 /bin/mail source: 554 lines.|     Henry Spencer at U of Toronto Zoology
1989 X.400 specs: 2200+ pages. | uunet!attcan!utzoo!henry henry@zoo.toronto.edu