Talk:Number One Electronic Switching System

NYTel
I worked for New York Telephone from January 1972 to Sept. 1982 before going over to Long Lines. I worked on one of the first #1ESS switches in Mahattan at the 240 East 38th Street Central Office. The day crew consisted of about 10 switchman who were assgined various tasks like watching the Master Control Center (MCC) and the output from the Tele-type Writer(TTY). New hires like myself went to work repairing inter-office trunks that connected our machine to the rest of the PSTN. As I got more experienced, I worked on various peripheral units like Line Switch frames. It was a great job and you were on your feet allot. I transfered to the CO at 435 West 50th Street in 1977 and in 1980/81 worked with Western Electric to replace the #1 proccesor with a 1A. Milepost53 (talk) 16:14, 8 August 2009 (UTC)
 * Ah the good old days when we had hair and hope and quick minds. Perhaps you were at the other end of a Disk File when we lifted it off the cart and slid it into the File Store on the 5th floor of NYCMNY50CG01, an office that mixed the older relay-tree SDs with the newer Remreed switches.  Jim.henderson (talk) 14:19, 10 August 2009 (UTC)

1ESS "Service Observing"
I worked for Chesapeake & Potomac Telephone Company of Virginia (became Bell Atlantic then Verizon) in Alexandria and Arlington from 1968 to 1979. I was fascinated by the new technology and even subscribed to the Bell System Technical Journal and Bell Labs Record to learn all I could. Started as a Frameman for #5 XBAR then worked as a Switchman on #1ESS, TSPS, and 4A-ETS/PBC at Arlington, Barcroft and Columbia Pike Central Offices. After being turned down for a programming job at C&P in Maryland, I left in 1979 to work for Digital Switch Corp. for 15 years on development of switching systems for the OCC providers such as MCI, Sprint, TeleSphere, Motorola Cellular, and DDI in Japan. Digital Switch became very successful and eventually was acquired by Alcatel and Lucent. Worked for Samsung for 7 years on 3G wireless core networks. Then went to work for EDS in Plano on Global Network Management for 4 years before they were acquired by HP and sent the jobs to the far east. Hold a BS from Univ of Texas Dallas and MS from SMU. Taught college part time for 8 years. There has been a lot of discussion in the news about "government surveillance" but this has been going on for many years. The 1ESS had special circuits connected to the "Number 2 Service Evaluation System" that was capable of collecting calling and called numbers (metadata) as well as monitor the voice path of calls. Special circuits were also provided to monitor individual subscriber lines. As part of my job at C&P I personally installed many wiretaps at the request of C&P management for "security and law enforcement representatives". Many operations systems developed at Bell Labs were described and photographs published in BSTJ and BLR. Look at the Bell System Practices of the day and you will find lots of information on the design, installation, maintenance, and repair of these systems. So then, what's all the fuss about NSA? The phone companies have captured this data for many decades. Personally, I am happy to see law enforcement use whatever tools they can to catch the bad guys. Telecom eng (talk) 07:41, 2 April 2014 (UTC)
 * Better to put the autobiography on your own WP:USERPAGE where anyone who sees any of the pages you have edited can see it. As they can see mine. Vaguely I recall doing something with SO near the end of my 5XB years, and I know it was in 1ESS, too. Only connection I had with wiretaps was an afternoon early in my 1ESS years, assigned to guide the tap gang of three around our large and oddly numbered MDF. As for putting Service Observing in a Wikipedia article, seems to me the 1ESS article would be an odd place. Well, maybe for a start, but perhaps plenty of information is available about history, purposes and methods. If so, it would quickly expand to be its own small article. Alternatively you could start the aticle on your own userpage or a subpage, and invite comment.  Jim.henderson (talk) 21:00, 3 April 2014 (UTC)

Non-technical human interest
This article would benefit from some non-technical information. For example, why was Succasunna chosen as the location of the first ESS? How did this affect normal subscribers? As this was the first opportunity for people to use call forwarding, call waiting and numerous other features, it might be good to have this angle mentioned in the article. Otherwise, to the normal reader it looks like just a bunch of technical jargon. YBG (talk) 12:46, 30 September 2015 (UTC)

Nonblocking minimal spanning switch
I just made a substantial edit resulting in the following sub-optimal block of text:

The traditional implementation of a nonblocking minimal spanning switch able to connect $$n$$ input customers to $$n$$ output customers simultaneously—with the connections initiated in any order—the connection matrix scaled on $$n^2$$. For example, for a switch with 1000 input customers and 1000 output customers, a traditional system able to route 1000 point-to-point connections simultaneously—in any possible permutation—would require a matrix of 1000×1000, or 1 million, physical switches for full interconnection possibility. When one considers that a large telephone system can have many more than 1000×1000 customers, the hardware to establish a full interconnection can grow rapidly and exceed practical implementations. Edson Erwin devised the Clos network in 1938 (as formalized by Charles Clos in 1952) which improved on the $$n^2$$ bound. Based on this conceptual foundation, Agner Krarup Erlang theorized a compromise which is based upon the concept that not all telephones lines are connected at the same time. From statistical theory, it is possible to design hardware that can connect most of the calls, in the sense of a high percentage, and block others as exceeding the design capacity. These are commonly referred to as blocking switches and are the most common in modern telephone exchanges. They are generally implemented as smaller switch fabrics in cascade. In many, a randomizer is used to select the start of a path through the multistage fabric so that the statistical properties predicted by the theory can be gained. In addition, if the control system is able to rearrange the routing of existing connections on the arrival of a new connection, a full non-blocking matrix requires fewer switch points.

The previous version implied that there was no known solution below n^2. But I happen to know about Clos networks as a concept in computer science, though I have only the barest fundamentals in telecommunications switching theory.

Note that nonblocking minimal spanning switch has a pretty good direct discussion of the 1ESS.

I think this exposition is still terribly clunky, but at least it now points at a correct conceptual model. Please wade in to fix all the small errors I must surely have made here. &mdash; MaxEnt 21:37, 29 May 2020 (UTC)

Call-back principle
The text uses this phrase, but I was unable to find any citation which defines this in a three-minute Google search with numerous permutations of the forced-item quotation marks. &mdash; MaxEnt 22:17, 29 May 2020 (UTC)

Failure → Benevolent Deception
Added a stub paragraph to the end of Fabric Error; I couldn't find much more and don't know if this should be broken out of that specific switching fabric section:

Anybody know more about this or how common this sort of behavior is now in other systems?

–ASiplas (talk) 17:40, 27 December 2020 (UTC)