The Telecom Digest for November 17, 2010
Volume 29 : Issue 310 : "text" Format
Messages in this Issue:
====== 28 years of TELECOM Digest -- Founded August 21, 1981 ======
Telecom and VOIP (Voice over Internet Protocol) Digest for the
Internet. All contents here are copyrighted by Patrick Townson and
the individual writers/correspondents. Articles may be used in other
journals or newsgroups, provided the writer's name and the Digest are
included in the fair use quote. By using -any name or email address-
included herein for -any- reason other than responding to an article
herein, you agree to pay a hundred dollars to the recipients of the
Addresses herein are not to be added to any mailing list, nor to be
sold or given away without explicit written consent. Chain letters,
viruses, porn, spam, and miscellaneous junk are definitely unwelcome.
We must fight spam for the same reason we fight crime: not because we
are naive enough to believe that we will ever stamp it out, but because
we do not want the kind of world that results when no one stands
against crime. Geoffrey Welsh
See the bottom of this issue for subscription and archive details
and the name of our lawyer, and other stuff of interest.
Date: Tue, 16 Nov 2010 06:51:41 +0000 (UTC)
From: email@example.com (Garrett Wollman)
Subject: Re: ISDN, was Why is T-1 24 Channels?
In article <firstname.lastname@example.org>,
John Levine <email@example.com> wrote:
>I dunno why they did that, Euro-ISDN was plug and play and is still
>reasonably popular. In the US ISDN is dead other than for PRI trunks
>to PBXes, carried over T-1, natch.
And as a major point-to-point transmission medium for broadcast audio,
both studio-to-studio and studio-to-transmitter. Most sports and
entertainment venues will have numerous ISDN circuits in them -- some
built and then torn down for specific events, and some permanent
installs for ongoing use. (It's not uncommon to see major-market
stations with three or four studio-transmitter links -- private
microwave, T-1, ISDN, and 15k equalized pairs -- as the potential for
revenue loss more than justifies the line rental, even if you keep two
B channels nailed up 1440 minutes a day.)
Garrett A. Wollman | What intellectual phenomenon can be older, or more oft
firstname.lastname@example.org| repeated, than the story of a large research program
Opinions not shared by| that impaled itself upon a false central assumption
my employers. | accepted by all practitioners? - S.J. Gould, 1993
Date: Tue, 16 Nov 2010 10:50:11 -0500
From: Fred Goldstein <fgoldstein.SeeSigSpambait@wn2.wn.net>
Subject: Re: Why is T-1 24 Channels?
On Mon, 15 Nov 2010 23:06:45 -0500, Bill Horne wrote,
>The D-1 bank used 8-bit samples: the low-order bit was robbed for
>signalling only when on-hook, i.e., when the circuit was in use, the
>8th bit was part of the PCM sample. Some call this "Seven and One-Half
>This caused, of course, an ambiguity in supervision, since the
>low-order bit had to be timed by the banks to check that it was in the
>on-hook state for several frames before the bank would assert on-hook
>for the channel.
Well, no; you had to use the low-order bit during the call to know
when it was over. I think the D1, which was pretty rare (basically
a prototype), robbed the bit in all frames. That's what I had in my
1991 book, ISDN In Perspective. and I'm sticking to my story. By the
time of the D3, which was mass-produced, they went to a sampling
system, with the bit robbed in every sixth frame. So the superframe
determined which samples had robbed bits and which didn't.
>BTW, I believe that robbed-bit signalling was the reason for ISDN's
>poor reputation during it's breif heyday as a data communications
>tool: even after Extended SuperFrame (ESF) removed signalling info
>from the individual channels' PCM stream,
You're confusing two things here. ESF didn't change the robbed-bit
signaling. ESF went to a longer superframe, allowing some of the
framing bits to be used for additional line-monitoring purposes. But
D4 channel banks still had robbed bits, and today, "channelized T1"
still emulates the D4.
Signaling was moved out of band in ISDN, via the D channel. In
Europe, however, they never had robbed bits; it was always 8-bit
clear. E1 carrier had 32 channels, one for timing and framing,
and one for signaling, leaving 30 voice channels. ISDN in Europe
meant changing the existing D channels from bit-oriented signaling to
message-oriented signaling. In the US, it meant moving the signaling
out of band, to a D channel, leaving 23 B channels, though you could
theoretically consolidate multiple PRIs and have the D channel on one
(with hot standby on a second) control a bunch of 24-channel
circuits. I don't think that is very popular though.
>the Baby Bells were
>reluctant to retire their gargantuan inventory of seven-and-one-half
>bit circuit packs, so ISDN users were usually limited to 56 Kbps
>connections during data calls (bonding excepted, of course). There had
>been some limited progress toward "End-to-end-eight-bit" trunk groups
>toward the middle-to-late Nineties, but since ISDN calls were billed
>per-minute, most consumers preferred to use modems and "local" access
>numbers, which resulted in developments that pushed speeds to the
>56Kbps range on dialup modems, thus pounding the last nail into ISDN's
ISDN specs recognized the existence of channels that were not 8-bit
clear, but the stated reason was not bit robbing. Rather, the old T1
transmission systems used AMI (alternate mark inversion) signaling
that required that at least 1/8 of bits be "1s", and that no more
than 15 consecutive "0s" be transmitted. This was okay with the D1
to D4 banks, because an active voice call never sent an all-0 pattern
(it was omitted from the PCM code), and IIRC an empty channel was
filled with 1s. But that couldn't transmit two consecutive "0" bytes.
By the time ISDN came out, though, they had figured out B8ZS
signaling. Bipolar 8-zero substitution sent a special pattern,
technically a bipolar violation (two consecutive marks in the same
polarity), in place of a span of 8 consecutive 0s. That made the
underlying carrier system clean for any bit pattern. However, there
were a lot of AMI (non-B8ZS) cards in the 4ESS and even some 5ESS
switches. So the trunks were not all 8-bit clear.
So ISDN allowed the user to request different bearer-channel
options. If the caller requested "64000 bps data" but there were not
clear trunks, the call would fail. The user could thus request
"56000 bps data" if necessary and ride over AMI trunks. Voice calls,
on the other hand, assumed that the other end could be bit-robbed
signaling, so only the high-order 7 bits were generally useful.
>Had the trunk network been 8-bit capable at the start of the PC
>revolution, ISDN might have gotten enough traction to survive.
Not quite. John Levine correctly raises the price issue:
>Not quite. Data calls were 8-bit, but the Bells, in their eternal
>quest to maximize short term revenue and drive customers away, charged
>more for data calls than for voice calls. Typically voice calls were
>unmetered, but data cost several cents per minute. Users, not being
>totally stupid, noticed that so long as they only used the high seven
>bits, they could send data using voice calls. Fred remembers more
>details than I do, but my recollection is that this was known as
>DOVBS, data over voice bearer service.
DOVBS (data over voice bearer service) was indeed quite common in
some parts of the US. I was a major advocate for it. BTW there was
no "voice" bearer service per se. Voice calls used either the "3.1
kHz audio" or "speech" bearer service. There was usually no
difference, but the former was nominally fax/modem clear while the
latter could theoretically have more audio processing (compression)
allowed. Both required A-law to mu-law conversion on transpudular
circuits, so DOVBS was domestic-only.
Whether or not you needed DOVBS depended on where you were. Most
NYNEX tariffs were, as John said, cheaper for voice; at least on
residential. So I used DOVBS on one of their extended-local-calling
plans to make 56000 bps calls to work. Data here was
measured-only. But in some states, voice and data bearer calls were
priced the same, so there was no DOVBS.
I actually started using DOVBS during an ISDN beta test, at NYNEX'
suggestion. Our company (DEC) had a nice PBX network, with the ISDN
receive modem at one site that had channelized T1 trunks but no ISDN
incoming trunks. So DOVBS worked through the PBX and across the
company's network, even from places where NYNEX had not yet installed
any ISDN data-bearer trunks.
>The main reason that ISDN died is that the telcos grossly overpriced
>it, but it also didn't help that the North American version was and is
>painful to configure, with the equipment at both ends having to be
>manually programmed, and it didn't work if the programming didn't
>precisely match. I dunno why they did that, Euro-ISDN was plug and
>play and is still reasonably popular. In the US ISDN is dead other
>than for PRI trunks to PBXes, carried over T-1, natch.
Indeed true. North American ISDN had a truly insane "feature" called
the SPID (service profile ID). This was literally a password. You
needed to program the terminal device (phone, modem) with the SPID of
the line or it would not connect at all. This was cooked up by some
Centrex weenies who wanted to support multiple telephone sets on the
same ISDN line -- a Basic Rate line could theoretically support up to
eight devices contending for the two B channels -- but they didn't
think it through well at all.
Plus the Bells mostly only wanted ISDN for Centrex feature
phones. They associated modems and ISDN with that newfangled
Internet thingie, which they HATED WITH A PURPLE PASSION. While they
couldn't identify modem users on POTS lines, they knew that ISDN data
was popular for Internet access. So they suppressed ISDN as a way to
show their hatred of all things Internet.
Not much has changed in that regard.
Fred Goldstein k1io fgoldstein "at" ionary.com
ionary Consulting http://www.ionary.com/
+1 617 795 2701
Date: Tue, 16 Nov 2010 19:40:52 EST
From: Wes Leatherock <email@example.com>
Subject: Re: History--computer based information operator terminal system
In a message dated 11/15/2010 9:02:31 PM Central Standard Time,
> Initial "1" wasn't used because of the possibility of dialing a
> false "1" when removing the receiver from the switchhook, and
> initial "0" was reserved for Operator.
Bells Labs made a study, I believe in the 1950s or so, to see how many
false "1s" actually occurred at the start of a csll. After amassing
several million examples of dialing, they found not a single case of
the false "1".
Date: Tue, 16 Nov 2010 19:55:51 EST
From: Wes Leatherock <firstname.lastname@example.org>
Subject: Re: early CATV, was: Bell System Technical Journal, 1922-1983
In a message dated 11/16/2010 12:08:38 AM Central Standard Time,
> Most of the AT&T microwave towers, buildings and sites are owned by
> American Tower, who is trying to lease space on them for other radio
> uses such as cellphone.
One tower, right next to a highway, is being used by an Oklahoma City
television statioon for their weather radar. This is the same tower
Southwestern Bell used to mount a temporary link to the University of
Oklahoma football stadium during football season when Southwestern
Bell and AT&T Long Lines carried tv signals over the towers.
The guys who did the work still had the title of "telephone
repeaterman," although their main work by then was maintaining the
Southwestern Bell (not AT&T then) microwave routes throughout the
TELECOM Digest is an electronic journal devoted mostly to telecom-
munications topics. It is circulated anywhere there is email, in
addition to Usenet, where it appears as the moderated newsgroup
TELECOM Digest is a not-for-profit, mostly non-commercial educational
service offered to the Internet by Bill Horne. All the contents
of the Digest are compilation-copyrighted. You may reprint articles in
some other media on an occasional basis, but please attribute my work
and that of the original author.
The Telecom Digest is moderated by Bill Horne.
Contact information: Bill Horne
43 Deerfield Road
Sharon MA 02067-2301
bill at horne dot net
Subscribe: email@example.com?body=subscribe telecom
Unsubscribe: firstname.lastname@example.org?body=unsubscribe telecom
This Digest is the oldest continuing e-journal about telecomm-
unications on the Internet, having been founded in August, 1981 and
published continuously since then. Our archives are available for
your review/research. We believe we are the oldest e-zine/mailing list
on the internet in any category!
URL information: http://telecom-digest.org
Copyright (C) 2009 TELECOM Digest. All rights reserved.
Our attorney is Bill Levant, of Blue Bell, PA.
Finally, the Digest is funded by gifts from generous readers such as
yourself who provide funding in amounts deemed appropriate. Your help
is important and appreciated. A suggested donation of fifty dollars
per year per reader is considered appropriate. See our address above.
Please make at least a single donation to cover the cost of processing
your name to the mailing list.
All opinions expressed herein are deemed to be those of the
author. Any organizations listed are for identification purposes only
and messages should not be considered any official expression by the
End of The Telecom Digest (4 messages)