1AESS: does it run the trunk-side DS0s thru a channel bank?
Aaron Leonard
into the network fabric? Or do they always run thru a channel bank?
What I'm trying to figure out is whether 1A-attached server-side PCM
modems can get a PCM connect to a client modem
PCM modems. I.e. you would have:
[ 1 ][?]===T1===[5ESS]--POTS--[ modem ]
[ A ]
[ E ]
[ S ]
[ PCM modem bank ] ===trunk-side T1===[?][ S ]---POTS line---[ PCM client modem ]
If I have this right, then PCM should work thru the 1A *to a client on the
same 1A* as there would be only one A/D conversion between the two modems.
But the question is whether you could get PCM to the modem off of the 5E
in the picture. If the DS0 containing the PCM words from the PCM modem
bank can be digitally switched to the T1 going to the 5E, then this should
work. But if the 1A runs the DS0 thru a codec and then thru another
codec to get to the 5E, then the 5E-attached modem will max out at V.34.
I guess from the description that I've seen of the 1AESS is that it
has an "analog switching fabric with digital control", the prospects
are not good that it should support PCM to a different CO, but I'm
hoping that someone who's actually dealt directly with these beasts
can tell me something different ...
Cheers,
Aaron
Ed Schulz
>
> If you have a trunk-side T1 into a 1AESS, do those DS0s go out digitally
> into the network fabric? Or do they always run thru a channel bank?
>
> What I'm trying to figure out is whether 1A-attached server-side PCM
> modems can get a PCM connect to a client modem
> PCM modems. I.e. you would have:
>
> [ 1 ][?]===T1===[5ESS]--POTS--[ modem ]
> [ A ]
> [ E ]
> [ S ]
> [ PCM modem bank ] ===trunk-side T1===[?][ S ]---POTS line---[ PCM client modem ]
The [?] in each case is a DCT unit: Digital Carrier Trunk. It's
essentially a D4 Channel Bank. The analog side connects to the analog
switch fabric, while the digitial side is the trunk. I'm told that
sometimes separate channel banks are used, but the picture is
functionally the same.
My home is on a 1AESS and I get PCM connections locally and remotely.
The only interesting part is for the analog modem to figure out the
precise nature of the DCT codec. This is left as an exercise for our
competitors!
--
Ed Schulz
edsc...@lucent.com
Hooda Gest
Aaron Leonard wrote in message
<35eefa50....@news-sj-2.cisco.com>...
>If you have a trunk-side T1 into a 1AESS, do those DS0s go out
digitally
>into the network fabric? Or do they always run thru a channel bank?
1A 1ESS has an analog network. It is a very old machine.
If the customer comes into the machine as analog from his home then it
is possible to get a PCM connection because of the single A/D conversion
as the call goes out. If the customer is on a SLC being fed into the 1A
1ESS then he will have multiple A/Ds.
..
Jeff McClure
>
>
>1A 1ESS has an analog network. It is a very old machine.
>
>If the customer comes into the machine as analog from his home then it
>is possible to get a PCM connection because of the single A/D conversion
>as the call goes out. If the customer is on a SLC being fed into the 1A
>1ESS then he will have multiple A/Ds.
>
Also...let's say that the ISP is connected to that 1A switch through a T1.
As stated above, someone that has a pure analog line into that switch can
get PCM connections. But...a user from another wire center will almost
certainly _not_ get PCM. Let's follow the connection in that situation. The
line from the ISP comes into the CO, goes analog to go into the switch,
comes out of the switch, and then gets converted _back_ to digital as it
goes onto the interoffice trunk. It gets to the user's CO, and is eventually
converted to analog yet again before it reaches him. 2 D/A conversions in
the path. No dice.
--Jeff
Hooda Gest
Ed Schulz wrote in message <35EF0E4D...@lucent.com>...
>Aaron Leonard wrote:
>>
>> If you have a trunk-side T1 into a 1AESS, do those DS0s go out
digitally
>> into the network fabric? Or do they always run thru a channel bank?
>>
>>
[ 1 ][?]===T1===[5ESS]--POTS--[ modem ]
>> [ A ]
>> [ E ]
>> [ S ]
>> [ PCM modem bank ] ===trunk-side T1===[?][ S ]---POTS line---[ PCM
client modem ]
>
>The [?] in each case is a DCT unit: Digital Carrier Trunk. It's
>essentially a D4 Channel Bank. The analog side connects to the analog
>switch fabric, while the digitial side is the trunk. I'm told that
>sometimes separate channel banks are used, but the picture is
>functionally the same.
In not all cases will a customer come in through a T1 line. As he
describes it, tere will be a channel bank or some direct conversion to
analog from the customer carrier so that the call can be put through the
the 1ESS's analog network.
>My home is on a 1AESS and I get PCM connections locally and remotely.
>The only interesting part is for the analog modem to figure out the
>precise nature of the DCT codec. This is left as an exercise for our
>competitors!
But do you know for a fact that you get to your 1A 1ESS office via a
digital carrier? Because if you are analog all the way in, you will
only have a single A/D conversion on the call.
..
Ed Schulz
>Also...let's say that the ISP is connected to that 1A switch through a T1.
>As stated above, someone that has a pure analog line into that switch can
>get PCM connections. But...a user from another wire center will almost
>certainly _not_ get PCM
Jeff is right. I didn't answer the question that Aaron asked.
An ISP would not want to host on an analog switch, because only subscribers with direct copper
connections on the same switch could get > V.34 rates. Never mind ISDN.
--
Ed Schulz
edsc...@lucent.com
Hooda Gest
Ed Schulz wrote in message <6snat8$l...@bgtnsc02.worldnet.att.net>...
>Jeff McClure wrote in message ...
[trimmed]
>Jeff is right. I didn't answer the question that Aaron asked.
>
>An ISP would not want to host on an analog switch, because only
subscribers with direct copper
>connections on the same switch could get > V.34 rates. Never mind
ISDN.
I don't know that an ISP would be able to get ISDN on a 1A 1ESS switch,
much less the type of digital service a decent sized ISP would require.
Heck, I can't even get ISDN on my local switch and it is a DMS-100
(fully digital), they'd have to hook me up to a 5ESS some 20 miles away.
..
Tony Pelliccio
Just out of curiosity, are you in old Nynex territory now owned by Bell
Titanic? I know that to cheapen things, Nynex ordered DMS-100's without
alot of the features that came standard on the 5E.
Bell had lots of catching up to do here in Rhode Island since the entire
state is on digital switching and most of those are DMS-100's.
Tony
Stanley Cline
> I don't know that an ISP would be able to get ISDN on a 1A 1ESS switch,
As they are analog, 1A's don't support ISDN.
They can do virtually everything else, including trunk-side (although
not fully digital, since they have to be broken down to analog at the
CO end) T1's, Caller ID/CLASS services, etc. etc.
> Heck, I can't even get ISDN on my local switch and it is a DMS-100
> (fully digital), they'd have to hook me up to a 5ESS some 20 miles away.
Maybe they don't want to provide ISDN out of a DMS, or the DMS doesn't
support ISDN (they would rather backhaul to a 5E than buy ISDN-capable
generics and switching units.)
-SC
Hooda Gest
Tony Pelliccio wrote in message ...
[trimmed]
>>
>> I don't know that an ISP would be able to get ISDN on a 1A 1ESS
switch,
>> much less the type of digital service a decent sized ISP would
require.
>> Heck, I can't even get ISDN on my local switch and it is a DMS-100
>> (fully digital), they'd have to hook me up to a 5ESS some 20 miles
away.
>
>Just out of curiosity, are you in old Nynex territory now owned by Bell
>Titanic? I know that to cheapen things, Nynex ordered DMS-100's without
>alot of the features that came standard on the 5E.
Nope. Down in south Florida... a wholly owned subsidiary of Bellsouth..
:)
The DMS-100 is a very popular switch throughout the country. It is
cheaper than the 5E and is apparently quite satisfactory for residential
use and most business use. Cost effectiveness is the primary criteria in
any business and the telcos are definitely ones who adhere to that
philosophy.
I do hear things from those who work in local switching that the DMS's
are not as easy to work with but no telco ever really listens to the
people who actually work on the equipment.
..
Tony Pelliccio
> The DMS-100 is a very popular switch throughout the country. It is
> cheaper than the 5E and is apparently quite satisfactory for residential
> use and most business use. Cost effectiveness is the primary criteria in
> any business and the telcos are definitely ones who adhere to that
> philosophy.
Actually my office is served by Brooks Fiber - they're using a 5E/2000.
Thank God some telcos know quality gear.
Tony
Art Jackson
>
> On Fri, 4 Sep 1998, Hooda Gest wrote:
>
>
>
> They can do virtually everything else, including trunk-side (although
> not fully digital, since they have to be broken down to analog at the
> CO end) T1's, Caller ID/CLASS services, etc. etc.
>
> > (fully digital), they'd have to hook me up to a 5ESS some 20 miles away.
>
> support ISDN (they would rather backhaul to a 5E than buy ISDN-capable
> generics and switching units.)
>
> -SC
The 1AESS office here was converted to a hybrid type with a DMS-100
attached to do trunk switching, and it also provided ISDN service,
although there was little demand for it...Several of those hybrid
1A's were in service around KY for a while, but have all been
replaced with fully digital switches now...We got a Seimens EWSD,
but it works just fine...Now if they would just replace those old
Universal SLC's, there would be a lot more happy campers here...
--
*************************** _______
| Art Jackson W4TOY | When all else | |
| Owensboro, Kentucky USA | fails. Read THE | HOLY |
| artj...@mindspring.com | Instruction Book. | BIBLE |
*************************** |_______|
Floyd Davidson
The DMS series of digital switches is just as much quality gear as are
the 4E and 5E switching systems. There are few things that either of
them will do that the other won't. The hardware design and implementation
is also of equal quality. It is a fact that DMS craft level software
is more versatile; however, that should not be construed as necessarily
better either since many telco's simply do not allow craft to access or
use even a small percentage of the capability.
AT&T, for example, operates both a 4E and a DMS-200 in
Anchorage, Alaska. Suffice it to say the 4E is there because
AT&T has a 4E network, and a DMS alone will just not interface
properly. The DMS-200 is there to do things a 4E can't do.
Floyd
--
Floyd L. Davidson fl...@ptialaska.net
Ukpeagvik (Barrow, Alaska)
Hooda Gest
Floyd Davidson wrote in message <6sphtv$7...@enews3.newsguy.com>...
[trimmed]
>
>AT&T, for example, operates both a 4E and a DMS-200 in
>Anchorage, Alaska. Suffice it to say the 4E is there because
>AT&T has a 4E network, and a DMS alone will just not interface
>properly. The DMS-200 is there to do things a 4E can't do.
Educate me... What can a DMS200 do that a 4ESS cannot do?
I have seen DMS-200's interfaced with AT&T's network, they have to be.
AT&T connects into DMS-200's as telco tandem switches all the time. If
the DMS can do SS7 signaling, it will function within AT&T's network.
..
Floyd Davidson
Note that I said "suffice it to say", which means that I am at
liberty to mention the fact that there are those two switching
systems in Anchorage... but my job description does not include
formulating public statements of exactly what they do or don't
do, which is company private information. Perhaps I can say
that it isn't some specific feature package that is or isn't
included, but trunk configuration and routing options that
are unique to the Alaska portion of the AT&T network. That
portion of the network was implemented before AT&T owned the
Alaska operation (pre-1995).
A DMS-200 can certainly be interfaced as a non-AT&T gateway to
the AT&T network. The AT&T network itself is based on the 4E
and integrating a DMS-200 as a part of that network certainly
could be done, but is simply not reasonable. Out of the 137 or
138 (I don't know what the count is now) tandem switches in the
AT&T network, the last I knew the DMS-200 in Anchorage was the
*only* DMS-200. That is hardly earth shaking, as very few
companies mix and match the two, and most that have DMS systems
do not have any 5E or 4E systems, and visa versa.
Hooda Gest
Floyd Davidson wrote in message <6sqjhd$m...@enews2.newsguy.com>...
>>
>>I have seen DMS-200's interfaced with AT&T's network, they have to be.
>>AT&T connects into DMS-200's as telco tandem switches all the time. If
>>the DMS can do SS7 signaling, it will function within AT&T's network.
>
>Note that I said "suffice it to say", which means that I am at
>liberty to mention the fact that there are those two switching
>systems in Anchorage... but my job description does not include
>formulating public statements of exactly what they do or don't
>do, which is company private information. Perhaps I can say
>that it isn't some specific feature package that is or isn't
>included, but trunk configuration and routing options that
>are unique to the Alaska portion of the AT&T network. That
>portion of the network was implemented before AT&T owned the
>Alaska operation (pre-1995).
I am sure there are a few features that each has that the other does not
do quite so well but I would bet that what can do, the other also can do
as well as is needed.
I suspect that AT&T kept the machine because it makes economical sense
and not because it could do something the 4E could not.
On the other hand, the 4E is a slightly older machine and AT&T has
considered phasing it out in the past so it's possible they are using
this configuration for study. You do recall that the 4ESS was MFD at one
point in time?
>A DMS-200 can certainly be interfaced as a non-AT&T gateway to
>the AT&T network. The AT&T network itself is based on the 4E
>and integrating a DMS-200 as a part of that network certainly
>could be done, but is simply not reasonable. Out of the 137 or
>138 (I don't know what the count is now) tandem switches in the
>AT&T network, the last I knew the DMS-200 in Anchorage was the
>*only* DMS-200. That is hardly earth shaking, as very few
>companies mix and match the two, and most that have DMS systems
>do not have any 5E or 4E systems, and visa versa.
Interfacing is now done at a 45Mbit level for physical connections
(through DACS) and via a standard signaling so there is nothing
inherently difficult about interfacing the machines. Timing might be an
issue but that should be no real difficulty to overcome. Heck, I see 30
channel digital carrier (2 Mbit?) interfaced into 45Mbit though I don't
see it being done real efficiently.
As for the mix and match among telcos, it is quite common. The local
telco here has a mix of DMS and AT&T for local switching and use DMS-200
for tandem switches (the 4E was too expensive for telco needs when it
went in).
Someone else mentioned interconnecting a DMS and 1A-1ESS and also having
a Seimens switch. The telcos, at least down here in the lower 48, are
getting pretty flexible. Since AT&T and Lucent have parted company, you
may even see that happening in AT&T in the coming years. Especially now
that the CEO is not a product of the Bell System... I suspect that MCI
is a real hodgepodge already.
Floyd Davidson
>Floyd Davidson wrote in message <6sqjhd$m...@enews2.newsguy.com>...
>>>Educate me... What can a DMS200 do that a 4ESS cannot do?
>>>
>>>I have seen DMS-200's interfaced with AT&T's network, they have to be.
>>>AT&T connects into DMS-200's as telco tandem switches all the time. If
>>>the DMS can do SS7 signaling, it will function within AT&T's network.
>>
>>Note that I said "suffice it to say", which means that I am at
>>liberty to mention the fact that there are those two switching
>>systems in Anchorage... but my job description does not include
>>formulating public statements of exactly what they do or don't
>>do, which is company private information. Perhaps I can say
>>that it isn't some specific feature package that is or isn't
>>included, but trunk configuration and routing options that
>>are unique to the Alaska portion of the AT&T network. That
>>portion of the network was implemented before AT&T owned the
>>Alaska operation (pre-1995).
>
>
>I am sure there are a few features that each has that the other does not
>do quite so well but I would bet that what can do, the other also can do
>as well as is needed.
Well, that is about exactly what I said to start with, so I'm
not sure your repetition has a great deal of value.
>I suspect that AT&T kept the machine because it makes economical sense
>and not because it could do something the 4E could not.
What I'm describing is not what I suspect might be the case, it
is what we _are_ doing with the two switching systems. I'm not
going to provide any details, but rest assured that 1) in its
draft form, the initial plan was to turn down the DMS-200 when
the 4E came on line, and 2) if that were possible it would
happen immediately. There is nothing economical about running
two tandem switches in the same building when the served
population is only 650,000 people! (Literally, either switch
in its minimal configuration could handle all of Alaska and
have a 10% load.)
>On the other hand, the 4E is a slightly older machine and AT&T has
>considered phasing it out in the past so it's possible they are using
>this configuration for study. You do recall that the 4ESS was MFD at one
>point in time?
Eh? I'm not sure what you mean by that. Both the DMS and the
4E date back to the late 1970's, though the DMS was the first
switch with a fully digital switching fabric, and therefore is
slightly older than the 4E in that respect. As to the specific
pair of switches in Anchorage, the DMS is much older (early
1980's) compared to the 4E (installed in 1996). At $30+ million
per at a minimum, with close to 140 of them, I doubt that AT&T
is going to phase out the 4E. It might not look the same in
10-20 years as they are modified, but...
>>A DMS-200 can certainly be interfaced as a non-AT&T gateway to
>>the AT&T network. The AT&T network itself is based on the 4E
>>and integrating a DMS-200 as a part of that network certainly
>>could be done, but is simply not reasonable. Out of the 137 or
>>138 (I don't know what the count is now) tandem switches in the
>>AT&T network, the last I knew the DMS-200 in Anchorage was the
>>*only* DMS-200. That is hardly earth shaking, as very few
>>companies mix and match the two, and most that have DMS systems
>>do not have any 5E or 4E systems, and visa versa.
>
>
>Interfacing is now done at a 45Mbit level for physical connections
>(through DACS) and via a standard signaling so there is nothing
>inherently difficult about interfacing the machines. Timing might be an
>issue but that should be no real difficulty to overcome. Heck, I see 30
>channel digital carrier (2 Mbit?) interfaced into 45Mbit though I don't
>see it being done real efficiently.
That is only the hardware. There is one heck of a lot more to
the PSTN that hardware interfacing. And note that a great deal
of DS1 interfacing (as well as DS3) that is external to a DACS
still exists. DS3's into a DACS is certainly the standard
though.
>As for the mix and match among telcos, it is quite common. The local
Not among LD providers with any significant amount of network. The
network administration, and the software interfacing between them,
is just plain expensive. That is of overriding significance if the
initial selection was either the 4E or the DMS-200; however, an
initial choice of almost any other system and then a move towards
one of those when expansion is necessary is reasonable.
Also note that the software licensing policies of companies such
as Nortel is designed to make it economical to use only one
company's switching system. In essense, the software license
for a main CPU is exceedingly expensive, while otherwise
complete switches that run remotely from a different main CPU
are greatly less expensive. Hence, a single DMS-100/200 with
the expensive license is commonly used to control multiple
remote units that are virtually a full switching system minus a
licensed software load in residence. The cost reduction is more
than a million dollars per remote unit per operating system
upgrade!
>telco here has a mix of DMS and AT&T for local switching and use DMS-200
>for tandem switches (the 4E was too expensive for telco needs when it
>went in).
Case in point, for tandem switching, one type.
>Someone else mentioned interconnecting a DMS and 1A-1ESS and also having
>a Seimens switch. The telcos, at least down here in the lower 48, are
>getting pretty flexible. Since AT&T and Lucent have parted company, you
>may even see that happening in AT&T in the coming years. Especially now
>that the CEO is not a product of the Bell System... I suspect that MCI
>is a real hodgepodge already.
I'd be amazed if MCI is, for the exact same reason that AT&T
isn't. The original choice was based as much as anything on the
fact that AT&T owned what is now Lucent. I doubt that MCI even
considered for a moment the idea of sharing the design
engineering of their network with AT&T's switch manufacturing
arm!!! (Which is why Lucent is so much more able to market its
products now that AT&T has no connection with it.)
John McHarry
wrote:
>Note that I said "suffice it to say", which means that I am at
>liberty to mention the fact that there are those two switching
>systems in Anchorage... but my job description does not include
>formulating public statements of exactly what they do or don't
>do, which is company private information. Perhaps I can say
>that it isn't some specific feature package that is or isn't
>included, but trunk configuration and routing options that
>are unique to the Alaska portion of the AT&T network.
To dispel some of the mystery, long distance networks often deviate
from standard SS7 signaling. These are usually proprietary extensions
to implement enhanced functions, like the ability to bill odd call
types from one record. It isn't surprising that there would be some
things a DMS could do in a given network that a 4E could not.
Hooda Gest
Floyd Davidson wrote in message <6ss621$p...@enews4.newsguy.com>...
>Hooda Gest <be@one_with.com> wrote:
>>
>>I am sure there are a few features that each has that the other does
not
>>do quite so well but I would bet that what can do, the other also can
do
>>as well as is needed.
>
>Well, that is about exactly what I said to start with, so I'm
>not sure your repetition has a great deal of value.
Basically, it was an agreement with that portion.
>>I suspect that AT&T kept the machine because it makes economical sense
>>and not because it could do something the 4E could not.
>
>What I'm describing is not what I suspect might be the case, it
>is what we _are_ doing with the two switching systems. I'm not
>going to provide any details, but rest assured that 1) in its
>draft form, the initial plan was to turn down the DMS-200 when
>the 4E came on line, and 2) if that were possible it would
>happen immediately. There is nothing economical about running
>two tandem switches in the same building when the served
>population is only 650,000 people! (Literally, either switch
>in its minimal configuration could handle all of Alaska and
>have a 10% load.)
Quite true about the capacity situation. The 4E in this town handles
that level of population with ease and is nowhere near a full machine.
>>On the other hand, the 4E is a slightly older machine and AT&T has
>>considered phasing it out in the past so it's possible they are using
>>this configuration for study. You do recall that the 4ESS was MFD at
one
>>point in time?
>
>Eh? I'm not sure what you mean by that. Both the DMS and the
>4E date back to the late 1970's, though the DMS was the first
>switch with a fully digital switching fabric, and therefore is
>slightly older than the 4E in that respect. As to the specific
>pair of switches in Anchorage, the DMS is much older (early
>1980's) compared to the 4E (installed in 1996). At $30+ million
>per at a minimum, with close to 140 of them, I doubt that AT&T
>is going to phase out the 4E. It might not look the same in
>10-20 years as they are modified, but...
I have no idea what Nortel's plans are regarding the DMS models so I
can't comment on that. AT&T's plans were public.. about 11 years ago,
they stopped manufacture (MFD) and were going to use the 5E in place (a
few were installed as LD switches). That didn't work out as well as
hoped and the 4E was started up again. There was a major modification a
few years ago that basically extended the life of the system.
The 4E has changed constantly over the years and to compare the first
machines ( a couple are still out there and functioning, a couple have
been retired) to the latest is apples and oranges. The first 4E went on
the PSTN in 1974 or 1975. Design was probably begun about 1970.
I don't know what kind of changes the DMS has gone through over the last
23 years.
>>Interfacing is now done at a 45Mbit level for physical connections
>>(through DACS) and via a standard signaling so there is nothing
>>inherently difficult about interfacing the machines. Timing might be
an
>>issue but that should be no real difficulty to overcome. Heck, I see
30
>>channel digital carrier (2 Mbit?) interfaced into 45Mbit though I
don't
>>see it being done real efficiently.
>
>That is only the hardware. There is one heck of a lot more to
>the PSTN that hardware interfacing. And note that a great deal
>of DS1 interfacing (as well as DS3) that is external to a DACS
>still exists. DS3's into a DACS is certainly the standard
>though.
The basic interface into the 4E (and I would assume the DMS) is at the
DS1 level although that is changing and it will be at the DS3 level soon
enough.
Timing and signaling are the primary issues. But, granted, features are
also important issues and integrating them could be much harder.
>>As for the mix and match among telcos, it is quite common. The local
>
>Not among LD providers with any significant amount of network. The
>network administration, and the software interfacing between them,
>is just plain expensive. That is of overriding significance if the
>initial selection was either the 4E or the DMS-200; however, an
>initial choice of almost any other system and then a move towards
>one of those when expansion is necessary is reasonable.
It is certainly always easier to add new switches of the same type than
to add switches from another manufacturer. However, GTE and Sprint have
done it in the past could well be doing it now.
>Also note that the software licensing policies of companies such
>as Nortel is designed to make it economical to use only one
>company's switching system. In essense, the software license
>for a main CPU is exceedingly expensive, while otherwise
>complete switches that run remotely from a different main CPU
>are greatly less expensive. Hence, a single DMS-100/200 with
>the expensive license is commonly used to control multiple
>remote units that are virtually a full switching system minus a
>licensed software load in residence. The cost reduction is more
>than a million dollars per remote unit per operating system
>upgrade!
Much the same as a 5ESS... Main processor, a number of remote units that
can handle fair size neighborhoods and some small towns.
The DMS-200 may be able to do both but I have not heard of one being
used in that manner. The 5E can be configured as one or the other but I
don't know that it can serve both functions simulataneously (it can also
function as a Operator Services unit but, again, not while also doing
tandem or local dialtone duty).
>>telco here has a mix of DMS and AT&T for local switching and use
DMS-200
>>for tandem switches (the 4E was too expensive for telco needs when it
>>went in).
>
>Case in point, for tandem switching, one type.
I understand that at least one telco wanted a bunch of 5Es for use as
local switches at Divestiture but WE was not prepared for the demand...
They switched to DMS100 but there was a scandal over that, according to
rumor.
>>Someone else mentioned interconnecting a DMS and 1A-1ESS and also
having
>>a Seimens switch. The telcos, at least down here in the lower 48, are
>>getting pretty flexible. Since AT&T and Lucent have parted company,
you
>>may even see that happening in AT&T in the coming years. Especially
now
>>that the CEO is not a product of the Bell System... I suspect that MCI
>>is a real hodgepodge already.
>
>I'd be amazed if MCI is, for the exact same reason that AT&T
>isn't. The original choice was based as much as anything on the
>fact that AT&T owned what is now Lucent. I doubt that MCI even
>considered for a moment the idea of sharing the design
>engineering of their network with AT&T's switch manufacturing
>arm!!! (Which is why Lucent is so much more able to market its
>products now that AT&T has no connection with it.)
I think MCI is already a hybrid system. I know their carrier systems
are. I see no real reason for them to do otherwise, if it satisfies
their needs, with their switching systems. Yes, it is easier to
integrate like switches but the manufacturer will bend over backward to
integrate its switches into your network; it's the smart thing to do
market-wise. AT&T and Nortel are certainly not the only makers of
telecommunications equipement in the world. If you aren't running into
some new (to AT&T) vendors up there in Alaska, best get prepared...
And you are definitely correct about Lucent's ability to market its
products now that they are not under AT&T's corporate umbrella.
Hooda Gest
Floyd Davidson wrote in message <6ssju2$c...@enews3.newsguy.com>...
>Hooda Gest <be@one_with.com> wrote:
>>
>>I have no idea what Nortel's plans are regarding the DMS models so I
>>can't comment on that. AT&T's plans were public.. about 11 years ago,
>
>been made public in the past are necessarily still active... :-)
I just mentioned the 11 year old plan where they discontinued the 4E and
then resurrected it. They did it again then brought in the 1B. They also
have some other plans that came and went and were replaced by still
another plan...
AT&T is very good at planning... :)
>We have shooed away two CEO's since then... :-) And divested huge
>amounts of both muscle and flab. (Now, if they would merely layoff
>90% of middle management, and 99% of upper management... we could
>get out of the Dilbert mode.)
One can only hope that the right people took that buyout recently...
THough it seems that the right ones never do... :(
>>The 4E has changed constantly over the years and to compare the first
>>machines ( a couple are still out there and functioning, a couple have
>>been retired) to the latest is apples and oranges. The first 4E went
on
>>the PSTN in 1974 or 1975. Design was probably begun about 1970.
>
>I'm not sure what the dates were. Simple fact is that Northern
>Telecom beat the pants off of AT&T with an early design of a
>digital switching system. The reason was because NTI built a
>switch assuming that an all digital switching fabric would be
>available, and when it was they built a new network module and
>dropped it in to the existing switch. AT&T had to come up with
>an entirely new design practically starting at the floor level.
>The result was that NTI grabbed a 40% market share out of what
>had been an almost 100% AT&T market, in a matter of about 5
>years. It was all a result of the decision to build an "as if"
>digital switch. It also happens that the decision was reputed
>to have come not from any committee or engineering group or
>whatever, but from the CEO himself. (My memory may be wrong,
>but I believe that was in 1975. I may be confusing the decsion
>date and the digital network module though.)
Interesting. The 3rd (I think) 4ESS was installed in Jacksonville, FL in
1975. Chicago was a year before that I believe. I do recall, however,
that around 1974 GTE claimed to be installing an all digital switch in
the Los Angeles area. I thought it was nothing much different that the
4A ETS at the time but more likely it was an early NT switch.
I don't think these two were that far apart in bringing the digital
network fabric out. Time Division networks were pretty much the path to
take at that point.
>>I don't know what kind of changes the DMS has gone through over the
last
>>23 years.
>
>Very similar to the 4E. The original racks and fuse panels will still
>function, and maybe a few power supplies and some of the maintenance
>modules. Very little else would be the same.
I have been in various offices over the years and the differences in
these old to new switches (not to mention local switches) are
incredible. Heck, they taught us about core stores and our machine came
in with SC. And that was in the space of a few months.
>>>>Interfacing is now done at a 45Mbit level for physical connections
>...
>>>That is only the hardware. There is one heck of a lot more to
>>>the PSTN that hardware interfacing. ...
>...
>>The basic interface into the 4E (and I would assume the DMS) is at the
>>DS1 level although that is changing and it will be at the DS3 level
soon
>>enough.
>>
>>Timing and signaling are the primary issues. But, granted, features
are
>>also important issues and integrating them could be much harder.
>
>Once again, the hardware interface (timing and signaling) are the
*easy*
>issues. Those are standardized. Configuration and routing options are
>an entirely different matter, and much more complicated. That has been
>made orders of magnitude more true with equal access. It is also more
>difficult because of variations in inter- and intra-state regulations.
Routing has changed over the years also. One upon a time, you went
through a straight hierarchy now you have one that changes constantly.
The advantages of SS7, I guess. Still, I think that the routing problems
are not so severe within your own netowork, it's interfacing with
someone els's network where the problems crop up. Within your own
network, you can control what call goes where and simply (through SS7)
control the route. It's been rather interesting as we integrate the
local telco switch network into ours... At least, that was done down
here.
>>It is certainly always easier to add new switches of the same type
than
>>to add switches from another manufacturer. However, GTE and Sprint
have
>>done it in the past could well be doing it now.
>
>When a company makes a gross error in selecting switching systems, it
>of course can be a good decision to make a change. Any company that
>chose either the Nortel or Lucent line has no such incentive, and
>virtually any other choice means they should be revisiting that
>choice at some point... but only one time. (That is an easy statement
>to make now, but in 1980 it was not at all that obvious.)
No one could know where the change in telecommunications would take us
in 1980. They really didn't even know in 1984 when Divestiture was
implemented. As I mentioned, a certain telco tried to buy a bunch of
5Es, WE Co could not deliver since they didn't foresee the demand. The
result was a huge boon to NT who made grandiose promises (on which they
failed to deliver immediately) about features and capabilities.
I had a vague idea in 1977 that our 13 man crew would be reduced to
almost nothing within a few years. I was wrong, it took longer than I
thought but I didn't count on the tenacity of lower management to
justify the unjustifiable.
>
>>Much the same as a 5ESS... Main processor, a number of remote units
that
>>can handle fair size neighborhoods and some small towns.
>>
>>The DMS-200 may be able to do both but I have not heard of one being
>>used in that manner. The 5E can be configured as one or the other but
I
>>don't know that it can serve both functions simulataneously (it can
also
>>function as a Operator Services unit but, again, not while also doing
>>tandem or local dialtone duty).
>
>The DMS switches are commonly configured as a DMS-100/200,
>meaning they do both lines and trunks. The only instance that I
You mean they can work as a tandem and as dialtone switch at the same
time? Interesting. I am not speaking of just normal inter-office
crosstown trunking but being a gateway into an LD network for multiple
COs while still acting as a CO for the folks in their immediate area. Am
I misunderstanding you?
Had AT&T developed a switch that way, they might have had a better shot
at entering the local market.
>>
>>I think MCI is already a hybrid system. I know their carrier systems
>>are. I see no real reason for them to do otherwise, if it satisfies
>
>There is a _huge_ difference between using a variety of carrier
>systems and a variety of switching systems. The best carrier system
>at any given time is the lowest priced one that will do whatever you
>are building. There is almost no penalty in having 14 different kinds.
>(I've got *six* different kinds in the AT&T Earth Station here in
>Barrow! And a couple other odd kinds at other locations along the
>Arctic Coast.)
There's a definite penalty in trying to mate 2Mbit systems into a system
that is designed to interface with 1.544 Mbit systems. It's not a pretty
sight... :)
Still, it can be done. As for integrating the switches, maybe I am
missing something but they connect via DS1 and routing can be commonly
controlled through STPs using SS7. Since SS7 is a standard, it's a
matter of understanding the way the individual switch handles an SS7
call. Now, where SS7 isn't implemented (and you may have more of that
than I have contact with) it would be much tougher. Similar to our
problems interfacing with GTE in the days prior to common channel
sinaling.
>>market-wise. AT&T and Nortel are certainly not the only makers of
>>telecommunications equipement in the world. If you aren't running into
>
>They are the only manufacturers of tandem switching systems which are
>even in the running.
That's true, I suppose. I was thinking more of all switching systems
combined. There is Ericcson (sp?), Seimens, and a Japanese one that
slips my mind. Telephonica in Mexico used to use a real strange mix, I
understand. Don't know if they still do.
We use a mix of DACS now, including Alcatel.
>>some new (to AT&T) vendors up there in Alaska, best get prepared...
>
>Giggle giggle. We are, by way far, the most diversified part of the
>AT&T network. We didn't have the legacy of the Bell System here,
>so neither the LEC's nor the IXC's have _ever_ tended to have the
>same habits that AT&T and the baby Bell's did. Our only Bell
>equipment was the original systems installed in the 1950's under
>various military contracts by Western Electric (the DEW Line and White
>Alice Comm System, both USAF systems, and the Alaska Communications
>System owned by the US Army. Everything after that was non-AT&T,
>and definitely represents the most diverse variety of telecom equipment
>in the US PSTN. (And trust me, AT&T has been having fits since
>they bought it in 1995 just trying to _inventory_ it!)
Then maybe it's me that ought to get prepared... I am still trying to
sort out the things being plugged in and keeping track of who makes what
and who I call for support.
>>And you are definitely correct about Lucent's ability to market its
>>products now that they are not under AT&T's corporate umbrella.
>
>The only really smart move, besides leaving, that Bob Allen made.
>(At least in my not too humble opinion!)
The sentiment toward Allen appears to be universal. One wonders why it
took so many years to move him out. He made some really poor decisions
while in charge.
I met him once, years ago, while he was still a step down from CEO. I
had no idea who he was. He could have been Mr. Rogers (of Mr. Rogers'
Neighborhood) for all I knew... he certainly looked like him. It was
only after he spent 15 minutes sitting and watching us and then left
with hardly a word said that I found out who he was.
Also met Joe Nacchio once (now head of Qwest but then a VP at AT&T).
Strange guy, made some strange comments. Still reminds me of a used car
salesman...
I think, though, that Qwest may be a good investment long term if they
can keep him honest.
Well, we're probably boring the shorts off everyone....
Floyd Davidson
>Floyd Davidson wrote in message <6ss621$p...@enews4.newsguy.com>...
>>Hooda Gest <be@one_with.com> wrote:
>
>>>On the other hand, the 4E is a slightly older machine and AT&T has
>>>considered phasing it out in the past so it's possible they are using
>>>this configuration for study. You do recall that the 4ESS was MFD at
>one
>>>point in time?
>>
>>Eh? I'm not sure what you mean by that. Both the DMS and the
>>4E date back to the late 1970's, though the DMS was the first
>>switch with a fully digital switching fabric, and therefore is
>>slightly older than the 4E in that respect. As to the specific
>>pair of switches in Anchorage, the DMS is much older (early
>>1980's) compared to the 4E (installed in 1996). At $30+ million
>>per at a minimum, with close to 140 of them, I doubt that AT&T
>>is going to phase out the 4E. It might not look the same in
>>10-20 years as they are modified, but...
>
>
>I have no idea what Nortel's plans are regarding the DMS models so I
>can't comment on that. AT&T's plans were public.. about 11 years ago,
I would not want to encourage you to think that AT&T's plans that have
been made public in the past are necessarily still active... :-)
We have shooed away two CEO's since then... :-) And divested huge
amounts of both muscle and flab. (Now, if they would merely layoff
90% of middle management, and 99% of upper management... we could
get out of the Dilbert mode.)
>The 4E has changed constantly over the years and to compare the first
>machines ( a couple are still out there and functioning, a couple have
>been retired) to the latest is apples and oranges. The first 4E went on
>the PSTN in 1974 or 1975. Design was probably begun about 1970.
I'm not sure what the dates were. Simple fact is that Northern
Telecom beat the pants off of AT&T with an early design of a
digital switching system. The reason was because NTI built a
switch assuming that an all digital switching fabric would be
available, and when it was they built a new network module and
dropped it in to the existing switch. AT&T had to come up with
an entirely new design practically starting at the floor level.
The result was that NTI grabbed a 40% market share out of what
had been an almost 100% AT&T market, in a matter of about 5
years. It was all a result of the decision to build an "as if"
digital switch. It also happens that the decision was reputed
to have come not from any committee or engineering group or
whatever, but from the CEO himself. (My memory may be wrong,
but I believe that was in 1975. I may be confusing the decsion
date and the digital network module though.)
>I don't know what kind of changes the DMS has gone through over the last
>23 years.
Very similar to the 4E. The original racks and fuse panels will still
function, and maybe a few power supplies and some of the maintenance
modules. Very little else would be the same.
>>>Interfacing is now done at a 45Mbit level for physical connections
...
>>That is only the hardware. There is one heck of a lot more to
>>the PSTN that hardware interfacing. ...
...
>The basic interface into the 4E (and I would assume the DMS) is at the
>DS1 level although that is changing and it will be at the DS3 level soon
>enough.
>
>Timing and signaling are the primary issues. But, granted, features are
>also important issues and integrating them could be much harder.
Once again, the hardware interface (timing and signaling) are the *easy*
issues. Those are standardized. Configuration and routing options are
an entirely different matter, and much more complicated. That has been
made orders of magnitude more true with equal access. It is also more
difficult because of variations in inter- and intra-state regulations.
>>>As for the mix and match among telcos, it is quite common. The local
>>
>>Not among LD providers with any significant amount of network. The
>>network administration, and the software interfacing between them,
>>is just plain expensive. That is of overriding significance if the
>>initial selection was either the 4E or the DMS-200; however, an
>>initial choice of almost any other system and then a move towards
>>one of those when expansion is necessary is reasonable.
>
>
>It is certainly always easier to add new switches of the same type than
>to add switches from another manufacturer. However, GTE and Sprint have
>done it in the past could well be doing it now.
When a company makes a gross error in selecting switching systems, it
of course can be a good decision to make a change. Any company that
chose either the Nortel or Lucent line has no such incentive, and
virtually any other choice means they should be revisiting that
choice at some point... but only one time. (That is an easy statement
to make now, but in 1980 it was not at all that obvious.)
>>Also note that the software licensing policies of companies such
>>as Nortel is designed to make it economical to use only one
>>company's switching system. In essense, the software license
...
>Much the same as a 5ESS... Main processor, a number of remote units that
>can handle fair size neighborhoods and some small towns.
>
>The DMS-200 may be able to do both but I have not heard of one being
>used in that manner. The 5E can be configured as one or the other but I
>don't know that it can serve both functions simulataneously (it can also
>function as a Operator Services unit but, again, not while also doing
>tandem or local dialtone duty).
The DMS switches are commonly configured as a DMS-100/200,
meaning they do both lines and trunks. The only instance that I
know of where it has not worked well was the original DMS
installed in Fairbanks, AK. That one was a joint effort by
Alascom, Inc. (then owned by Pacific Telecom, Inc.) and
Fairbanks Municipal Utilities System. FMUS physically operated
the switch, and they configured the 100 side while Alascom
used/configured the 200 side. It happens that there is not
really a clear distinction between what is "100" and what is
"200". The results were absolutely hilarious! Every day the
FMUS folks would do their best to lock Alascom techs out of
various parts of the switch, and every morning one of the
Alascom techs would hack the system to allow them back in. Fate
took hold though, and a fire damaged the switch beyond repair.
The fix was that both companies installed their own DMS. The
FMUS switch, by the way, was a DMS-100/200 to start with, while
the original FRBN toll switch was a DMS-200 that was later
converted to a DMS-100/200 to provide FTS-2000 requirements
under contract to AT&T. (AT&T bought Alascom, Inc. in 1995 and
deactivated that switch the next year.)
>>>may even see that happening in AT&T in the coming years. Especially now
>>>that the CEO is not a product of the Bell System... I suspect that MCI
>>>is a real hodgepodge already.
>>
>>I'd be amazed if MCI is, for the exact same reason that AT&T
>>isn't. The original choice was based as much as anything on the
>>fact that AT&T owned what is now Lucent. I doubt that MCI even
>>considered for a moment the idea of sharing the design
>>engineering of their network with AT&T's switch manufacturing
>>arm!!! (Which is why Lucent is so much more able to market its
>>products now that AT&T has no connection with it.)
>
>
>I think MCI is already a hybrid system. I know their carrier systems
>are. I see no real reason for them to do otherwise, if it satisfies
There is a _huge_ difference between using a variety of carrier
systems and a variety of switching systems. The best carrier system
at any given time is the lowest priced one that will do whatever you
are building. There is almost no penalty in having 14 different kinds.
(I've got *six* different kinds in the AT&T Earth Station here in
Barrow! And a couple other odd kinds at other locations along the
Arctic Coast.)
>their needs, with their switching systems. Yes, it is easier to
>integrate like switches but the manufacturer will bend over backward to
>integrate its switches into your network; it's the smart thing to do
>market-wise. AT&T and Nortel are certainly not the only makers of
>telecommunications equipement in the world. If you aren't running into
They are the only manufacturers of tandem switching systems which are
even in the running.
>some new (to AT&T) vendors up there in Alaska, best get prepared...
Giggle giggle. We are, by way far, the most diversified part of the
AT&T network. We didn't have the legacy of the Bell System here,
so neither the LEC's nor the IXC's have _ever_ tended to have the
same habits that AT&T and the baby Bell's did. Our only Bell
equipment was the original systems installed in the 1950's under
various military contracts by Western Electric (the DEW Line and White
Alice Comm System, both USAF systems, and the Alaska Communications
System owned by the US Army. Everything after that was non-AT&T,
and definitely represents the most diverse variety of telecom equipment
in the US PSTN. (And trust me, AT&T has been having fits since
they bought it in 1995 just trying to _inventory_ it!)
>And you are definitely correct about Lucent's ability to market its
>products now that they are not under AT&T's corporate umbrella.
The only really smart move, besides leaving, that Bob Allen made.
(At least in my not too humble opinion!)
Floyd
Floyd Davidson
>Floyd Davidson wrote in message <6ssju2$c...@enews3.newsguy.com>...
>>Hooda Gest <be@one_with.com> wrote:
>>
>>The DMS switches are commonly configured as a DMS-100/200,
>>meaning they do both lines and trunks. The only instance that I
>
>You mean they can work as a tandem and as dialtone switch at the same
>time? Interesting. I am not speaking of just normal inter-office
>crosstown trunking but being a gateway into an LD network for multiple
>COs while still acting as a CO for the folks in their immediate area. Am
>I misunderstanding you?
Works like a charm. Lines are handled by one type of hardware
peripheral, and trunks by another. With the proper software
modules routing of trunks is available, and with the other
proper software modules routing of lines is available. At one
time there were at least *six* DMS tandem switches in Alaska
owned by three different companies (including the old military
autovohn system) that also switched a few lines, and at least
two primarily line switches that also switched a few trunks.
Eric Hildum
Hooda Gest wrote:
> That's true, I suppose. I was thinking more of all switching systems
> combined. There is Ericcson (sp?), Seimens, and a Japanese one that
> slips my mind. Telephonica in Mexico used to use a real strange mix, I
> understand. Don't know if they still do.
>
Would the Japanese switch you are thinking of be the D-70? If so, that
design is manufactured by all four major telecom suppliers in Japan under
contract to NTT. I have been told that it is one of, if not the best, ISDN
switches around (but that is about all it is able to do well).
--
---------------------------
Eric Hildum
Eric....@Japan.NCR.COM
Hooda Gest
Eric Hildum wrote in message <35F47C41...@Japan.NCR.COM>...
>
>
>Hooda Gest wrote:
>
>> That's true, I suppose. I was thinking more of all switching systems
>> combined. There is Ericcson (sp?), Seimens, and a Japanese one that
>> slips my mind. Telephonica in Mexico used to use a real strange mix,
I
>> understand. Don't know if they still do.
>>
>
I wasn't thinking of a specific switch, just the manufacturer.. NEC is
what slipped my mind... (easy enough to do, since my mind is usually
slippery and sloped downward....)
That was from a number of years ago so who knows who else might be
making switches these days?
The point, I guess, was that while Western Electric (now Lucent) and
Northern Telecom (Nortel) are the first two names anyone thinks of, they
aren't the only ones to make telco switch systems.
..
Al Varney
S.C. <roamer1@-removeme-pobox.com> wrote:
>On Thu, 03 Sep 1998 21:01:54 GMT, Aa...@Cisco.COM (Aaron Leonard)
>wrote:
>
>>If you have a trunk-side T1 into a 1AESS, do those DS0s go out digitally
>>into the network fabric? Or do they always run thru a channel bank?
>
>
>There is no choice; the DS0s have to be broken down to analog.
>that trunk situation just fine.)
Actually, the 1A ESS(tm) switch has a "digital trunk" option that supports
56K digital being switched through the analog fabric as an analog waveform.
(The analog fabric isn't limited to 4KHz transmission. The fabric even
supports a form of 4-wire tandem switching using FDM to separate the 2
transmission directions.) The "digital" trunks can only support 56K data
(no voice, no 64K), as in "Switched 56 Data" service. Very little demand for
the service, and thus few switches have the capability. And you can't connect
POTS callers to the trunks anyway, because they are VOICE (not 56K data).
These trunks supported a form of pre-ISDN digital line (PSDC access) on the
1A ESS switch.
>I generally would advise against a '56k' ISP having trunks from a #1A
>unless a) the local calling area consists mostly or entirely of one CO
>(the #1A), and b) there is simply no other choice. (Some ISPs that
>want to do 56k in #1A-heavy areas, such as Rome and Waycross, GA and
>New Orleans, find it easier to locate their POP in an adjacent or
>colocated digital CO [or a CLEC's] service area, or backhaul trunks
>from the other CO.)
If I remember correctly, Waycross had a bunch of SXS CDOs hanging off
of it (in the 1980s). These can introduce a lot of noise into a POTS
connection. Maybe they've been replaced by something that supports
Equal Access??
Note that '56k modems' can work in the opposite configuration, where
the ISP connects at a digital switch and the caller is a POTS subscriber
on the 1A ESS switch. The switch fabric can have a lot of impulse noise
(during heavy calling periods) and other analog items (loading coils, etc.)
can also limit speed and increase error rate.
Al Varney
Al Varney
Floyd Davidson <fl...@ptialaska.net> wrote:
>Hooda Gest <be@one_with.com> wrote:
>>Floyd Davidson wrote in message <6ss621$p...@enews4.newsguy.com>...
>>>Eh? I'm not sure what you mean by that. Both the DMS and the
>>>4E date back to the late 1970's, though the DMS was the first
>>>switch with a fully digital switching fabric, and therefore is
>>>slightly older than the 4E in that respect.
The definitive history of switching (Amos Joel's "Electronics, Computers
and Telephone Switching") tells a slightly different story. But he's
been wrong before (once, back in 1955, I'm told. :) ).
One firm reference point: The first 4ESS went into service 17 Jan 1976.
As for DMS systems, I find (in roughly reverse order):
First DMS-200 in AT&T network, March 1983 (Chico, CA)
First DMS-200, January 1979 (Ottawa, Canada)
First DMS-100, September, 1979 (Ottawa, Canada)
(First NEAX61 in USA, 5 May 1979 in Manteca, CA)
First DMS-10 BOC switch, January 1980 (Union, KY)
First DMS-10 BC Tel switch, October 1978 (Embrun, Canada, near Ottawa)
First DMS-10 switch, October, 1977 (Disney World, 400 lines)
(Stromberg Carlson cut-over the "System Century DCO" in
Richmond Hills, GA on 16 July 1977 (500 lines), for Continental Tel.)
First 4ESS with digital trunks, "Kansas City 2", 3 July 1976.
(Vidar cut-over a fully-digital local tandem switch in China Lake, CA
on 26 March 1976.)
First 4ESS, "Chicago 7", 17 Jan 1976.
First DMS1 (similar to a SLC-96 line multiplexor), April, 1977
First SL1 PBX, 1975
So, if you consider the fabric of the SL1 PBX to be "fully digital",
then the SL1 beat the 4ESS switch by a few months. But not a DMS.
>>The 4E has changed constantly over the years and to compare the first
>>machines ( a couple are still out there and functioning, a couple have
>>been retired) to the latest is apples and oranges. The first 4E went on
>>the PSTN in 1974 or 1975. Design was probably begun about 1970.
January, 1976, as above. Earle Vaughan became Head of a new
Toll ESS department in April 1968. He did not convince the Switching
Council that a time-division switch would be cheaper than the analog
alternatives until early 1970.
>I'm not sure what the dates were. Simple fact is that Northern
>Telecom beat the pants off of AT&T with an early design of a
>digital switching system. The reason was because NTI built a
>switch assuming that an all digital switching fabric would be
>available, ...
The fabric was already available on the SL1. What wasn't obvious
was that the fabric would be reliable enough to support a LARGE CO.
But the alternative was to continue with the SP1 mini-X-Bar fabric,
and no one in 1974 would have decided to go with mini-X-Bar fabrics
-- well, almost no one. (No. 3 ESS was just such a system, designed
in 1974, into service in 1976. :( )
I might add that concern for long-term large-fabric reliability was
one reason the 4ESS switch was designed with a duplicated fabric.
The DMS-10 duplicated the central control of the SL1, but not the fabric.
>.... It was all a result of the decision to build an "as if"
>digital switch. It also happens that the decision was reputed
>to have come not from any committee or engineering group or
>whatever, but from the CEO himself. (My memory may be wrong,
>but I believe that was in 1975. I may be confusing the decsion
>date and the digital network module though.)
Well, it was a task force (committee) of Northern Telecom, BNR and
Bell Canada engineers that in 1975 looked at their options for replacing
the SP1 analog switch in Canada, and decided the SL1 with some changes
could both meet their needs AND provide an opportunity for small EO
sales in the USA, where they were currently selling a small X-Bar switch.
I'm sure Mr. Chisholm (BNR President at the time) deserves credit for
pushing the final decision, committing lots of capital into a development
that likely was a loss until the early 1980s. But in the mid-1970s, no
lone CEO of a regulated industry made that kind of decision.
>>Much the same as a 5ESS... Main processor, a number of remote units that
>>can handle fair size neighborhoods and some small towns.
>>The DMS-200 may be able to do both but I have not heard of one being
>>used in that manner. The 5E can be configured as one or the other but I
>>don't know that it can serve both functions simulataneously (it can also
>>function as a Operator Services unit but, again, not while also doing
>>tandem or local dialtone duty).
Actually, the 5ESS-2000(rg tm) switch can function as a CO, tandem,
IXC switch, wireless MSC and Operator Services switch simultaneously.
Some back-to-back trunking is needed to make some features work properly,
since the IXC, Operator Services and wireless MSC aren't expecting to
interact directly with POTS and ISDN lines. The internal fabric and
most other hardware doesn't change with the application.
>The DMS switches are commonly configured as a DMS-100/200,
>meaning they do both lines and trunks.
>.... It happens that there is not
>really a clear distinction between what is "100" and what is
>"200". The results were absolutely hilarious!
The No. 1/1A ESS switches were available in a combined local/toll
arrangement. Beaumont, TX, was one I visited. There were painted lines
on the floor and roped-off areas to confine AT&T Long Lines folks to
"their part" of the switch (mostly some trunk switch networks and the trunk
test equipment dedicated to their trunk circuits). Trunk/call failures
associated with Long Lines circuits were printed on AT&T-specific terminals,
which had access only to their trunks. The Davenport, IA, tandem was
another example -- divestiture drove the Toll folks out of that one.
Al Varney
Eric Hildum
had shipped anything outside of Japan. When did it happen?
Hooda Gest wrote:
>
>
> I wasn't thinking of a specific switch, just the manufacturer.. NEC is
> what slipped my mind... (easy enough to do, since my mind is usually
> slippery and sloped downward....)
>
> That was from a number of years ago so who knows who else might be
> making switches these days?
>
> The point, I guess, was that while Western Electric (now Lucent) and
> Northern Telecom (Nortel) are the first two names anyone thinks of, they
> aren't the only ones to make telco switch systems.
>
> ..
--
---------------------------
Eric Hildum
Eric....@Japan.NCR.COM
Robert Nichols
Al Varney <var...@lucent.com> wrote:
:But the alternative was to continue with the SP1 mini-X-Bar fabric,
:and no one in 1974 would have decided to go with mini-X-Bar fabrics
:-- well, almost no one. (No. 3 ESS was just such a system, designed
:in 1974, into service in 1976. :( )
Just for the record, No. 3 ESS used a remreed switch network, not
mini-X-bar.
Trivia note: Some of the switch modules used in No. 3 ESS offices were
factory rejects from the 1A ESS frame assembly line. No. 3 ESS had a
beefier network pulser that could reliably operate some crosspoints that
would not work in 1A ESS frames. That network pulser underwent an
emergency redesign after the transistors in one of the early models
exploded during an office installation. Exploding TO-3 cans tend to
attract attention, especially when people are standing nearby.
--
Bob Nichols rnic...@interaccess.com
Finger rnic...@cluster.interaccess.com for PGP public key.
PGP public key 1024/9A9C7955
Key fingerprint = 2F E5 82 F8 5D 06 A2 59 20 65 44 68 87 EC A7 D7
Al Varney
Robert Nichols <rnic...@interaccess.com> wrote:
>In article <6t40si$l...@ssbunews.ih.lucent.com>,
>Al Varney <var...@lucent.com> wrote:
>:and no one in 1974 would have decided to go with mini-X-Bar fabrics
>:-- well, almost no one. (No. 3 ESS was just such a system, designed
>:in 1974, into service in 1976. :( )
>
>mini-X-bar.
You're right, of course. I should have said that no one in 1974 would
be considering mechanically-switched fabrics. It did make the No. 3
ESS a very compact installation (compared to other analog switches).
The first 1A ESS also went into service in 1976, using the same remreed
switch fabric, but then that wasn't a new switch fabric (re-used the
No. 1 ESS fabric).
Al Varney
Art Jackson
>
> In article <Ez0uxz.763...@interaccess.com>,
> Robert Nichols <rnic...@interaccess.com> wrote:
> >In article <6t40si$l...@ssbunews.ih.lucent.com>,
> >Al Varney <var...@lucent.com> wrote:
> >:and no one in 1974 would have decided to go with mini-X-Bar fabrics
> >:-- well, almost no one. (No. 3 ESS was just such a system, designed
> >:in 1974, into service in 1976. :( )
> >
> >mini-X-bar.
>
> You're right, of course. I should have said that no one in 1974 would
> be considering mechanically-switched fabrics. It did make the No. 3
> ESS a very compact installation (compared to other analog switches).
> The first 1A ESS also went into service in 1976, using the same remreed
> switch fabric, but then that wasn't a new switch fabric (re-used the
> No. 1 ESS fabric).
>
> Al Varney
I maintained 4 TSPS Remote Trunk Arrangement(RTA's) in Western KY.
They were installed in 1978...They used the mini-Xbar switches in
those, and they were a constant source of problems...They were
connected to #5 Xbar's when they were installed, but they were
replaced by 1AESS's in 79 and 80...
Al Varney
True. In fact, the number of defunct vendors is probably larger than
the number of current vendors.... I can think of ITT, Vidar, Nokia, Italtel,
Automatic Electric (although AGCS still survives), Plessey/GEC, Phillips,
Thomson (sorta survives in Alcatel), and East German/Soviet companies.
Is Kellogg alive (in Siemens) or dead?
The market today is Alcatel, LME (Ericsson), Siemens Stromberg Carlson,
Lucent Technologies, Nortel, NEC and Fujitsu. I've probably missed some;
should I count DSC? Tomorrow, who knows? Microsoft, Cisco, Bellcore and
Ascend would all love to be "public switch" vendors. The rule of thumb
20 years ago was that 6 vendors was all the market would support (if
governments didn't interfere with the market). None of the new entrants
owe anything to a particular country. Or to the ITU, for that matter.
Just a quick Japan-export survey:
Analog: NEC marketed ND10/ND20 "minireed" switches to the USA "independents"
in the late 1970s. Fujitsu sold "memoreed" switches to Hong Kong and
Singapore in 1978. These were mini-crossbar switches upgraded to operate
with sealed-contact fabrics.
Digital: NEC sold NEAX-61s to the USA from 1979 on, many for ISDN
"trials". Dozens of countries bought various variations of the NEAX-61.
Fujitsu sold FETEX150 systems to Singapore, Hong Kong, China, etc. USA
installations were primarily for "ISDN trials". Hitachi sold HDX10s to
Sri Lanka and a few other countries. Don't know if OKI was successful
with the KB270 outside Japan.
Al Varney - just my opinion
BRian m Tausend
Cortelco which was purchased and is now called Cortelco Systems.
Panasonic is a Japanese phones system though the USA version is manufactured
in the USA.
Brian M. Tausend
Al Varney wrote in message <6ue1lv$o...@ssbunews.ih.lucent.com>...
Terry Kennedy
> Panasonic is a Japanese phones system though the USA version is manufactured
> in the USA.
Hmmm. Most of my modern Panasonic stuff was assembled in Wales. The earlier
stuff (same models) was made in Japan - the Wales factory was started to get
around US "anti-dumping" tariffs on Japanese products.
Terry Kennedy Operations Manager, Academic Computing
te...@spcvxa.spc.edu St. Peter's College, Jersey City, NJ USA
+1 201 915 9381 (voice) +1 201 435-3662 (FAX)
Eric Hildum
about carrier equipment, not systems intended primarily for CPE...