Talk:Public switched telephone network

POTS
Moved POTS to a separate page because the services are provided at the end of the network.

POTS, or Plain Old Telephone Service, describes the services available on analog telephones. This is a subset of what is available on ISDN and mobile telephones.

post Procsperco (talk) 14:52, 7 June 2018 (UTC)

procsperco Procsperco (talk) 15:03, 7 June 2018 (UTC)

prospere csoungali Procsperco (talk) 15:05, 7 June 2018 (UTC)

RE: Capitalization Question
Shouldn't public switched telephone network be capitalized in the article? 144.118.58.229 06:59, 7 November 2007 (UTC)
 * See the 'Article titles' section of Manual of Style for details. 'Bauani'  •  Talk 2 me  11:40, 26 April 2008 (UTC)
 * Public Switched Telephone Network, like "Internet", is a proper noun referring specifically to a certain, albeit large, switched telephone network -- which is to say, one that is *not* AUTOVON/DSN, etc, etc. It should, in fact, carry upstyle capitalization for that reason. --Baylink (talk) 22:27, 9 September 2009 (UTC)


 * I more or less agree, but you're going to have trouble trying to convince others of that opinion though for example the majority of edu and ac.uk sites seem to capitalise. See for example Internet_capitalization_conventions for how the telephone network is used to support lowercase usage of "Internet": ...some things that are unique yet distributed, such as "the power grid", "the telephone network", and even "the sky", are not considered proper nouns, and are thus not capitalized. On the other hand, the article's name is more of a proper noun than "the telephone network" due to the nearly official name "public switched telephone network", which is however still only a descriptive name i.e. common noun as long as there's no main office or any kind of central organisation. In fact, "the Internet" is (or was) probably only capitalised to differentiate it from (generic) internets i.e. internetworks. And there are no other public switched telephone networks that are generic. --Espoo (talk) 17:13, 10 September 2009 (UTC)

Pre-Merger Concept
I added the hierarchy with links to the main articles. Luis F. Gonzalez 18:03, 17 November 2006 (UTC)

Ought each section be so small, or would the longer descriptions in office classification be more appropriate? Jim.henderson 21:02, 17 November 2006 (UTC)


 * I think the concept of PSTN office classification should start in this article. With each section being small. The overall hierachy section should point the office classification article, and each class should point to its respective class article. Each class article then should point to specific switches, 5ESS/DMS. No? Luis F. Gonzalez 21:42, 17 November 2006 (UTC)


 * The first three classes don't have much to say, hence should contain everything that any existing article (especially Office classification) says about them, apart from any material that may be so useless that it ought simply to die.


 * I agree, the first three don't have much meat and don't need their own articles. But we don't know if someone at some future point may have more information. Luis F. Gonzalez 06:48, 22 November 2006 (UTC)


 * Sections for classes 4 and 5 must necessarily be larger, essentially pasting in whatever is in their specific class articles today. Yes, all class sections should also point to specific hardware.


 * The PSTN article in my opinion should have minimal information on class, one paragraph for example. Each class article 4 and 5 in could then have one para on specific hardware which then point to deeper vendor implementations. Luis F. Gonzalez 06:48, 22 November 2006 (UTC)


 * Umm, but no, you're saying to keep the Class 4 and Class 5 articles out of this one, thus their sections here would not be much more than definitions and links. Yes, that certainly will keep this one from becoming long and ungainly.  Is this your reason for retaining the class articles?


 * Yes. Luis F. Gonzalez 06:48, 22 November 2006 (UTC)


 * Well then, that leaves our little project rather small. It's just a matter of merging Office classification into this one.  Yes, that's workable and leaves room for many additions in years to come, if someone wants to make extensive additions.
 * Jim.henderson 15:17, 22 November 2006 (UTC)


 * That would be all very well if the USA was the entire world but of course it is not. Here we have a perfectly relevant document (PSTN article) for those users not located in the US potentially rendered irrelevant for the merging with a specifically US centric document. Granted, the larger proportion of users are US based users but quite often it seems many articles presume a specifically US audience.

Well then, I am inclined to agree with our anonymous commentator. The mention here of the antique American classes should be pared, and not merged. I'd like to see an article on Tandem switch from a less parochial view. Right now it's just a redirect to a strictly American article. Anyway in recent week I've been doing other Wikipedia things, for example describing Panel swtich, 1XB switch, 5XB switch, 1ESS switch, Reed relay and other topics, and paying no attention to how these switched network articles ought to be divided. So, I hope our friend Louis will take good care of it. Jim.henderson 04:43, 18 January 2007 (UTC)

I added a couple of paragraphs here to clarify the historical context. As written, it still looks as if the hierarchy is meaningful, rather than an artifact of the mechanical-switching era. The PSTN continues to evolve, so describing it in 1970s terms without clearer context is confusing. User:isdnip 15:22, 7 September 2009 (UTC).

Past tense?
I noticed that section 3 -- U.S. Telephone Switch Hierarchy -- is written in the past tense ("exchanges were"). Does this mean the U.S. TSH is no longer in use? I would reccomend either moving this to the infinitive tense ("exchanges are"), or clarifying the fact that this system no longer exists by mentioning that fact within the section. --Action Jackson IV 04:48, 18 January 2007 (UTC)


 * The concept was important in the Bell System. In the 1950s and 60s routing was in large degree governed by this office hierarchy.  As the system became larger and more complex, and as computerization and common channel signalling allowed more complex routing methods, the hierarchy became less relevant to ordinary operations.  Whether the system was ever formally abolished, I don't know.  Similar classifications were set up in foreign countries, and an international classification scheme of "Centres de Transit" with London as the first CT1 in the 1970s and lesser levels as CT2 and CT3, but again I don't know whether any of these systems are still in use or were ever abolished.


 * Anyway it has been proposed that this section be cut down or abolished as belonging to only one nation (even if that nation had the majority of the world's telephones at the time) leaving Office classification to handle these questions as they pertain to the USA. Jim.henderson 02:35, 22 January 2007 (UTC)

Wartime Situations/World War
What if there's another "World War" or military conflict among more than one countries; how will that affect the International Gateways? Will it have some significant effect on our Internet?

88.105.55.192 13:17, 6 October 2007 (UTC)

Classes given too much signifiance
The Discussion of the AT&T switch class nomenclature shouldn't be given so much significance within the article. Only North America ever used this convention and it doesn't deserve such importance in an article about what a Public Switched Telephone Network (PSTN) is, (and yes it should be capitalised when used as it's an acronym). The article should be focusing on the technology and topology used within the network. The fact that it was divided in to classes in the USA should constitue a regional variance side note at most, much like in the UK they are called: RCUs (Remote Concentrator Unit), DLSU (Digital Local Switching Unit), DMSU's (Digital Main Switching Unit) or DISC (Digital International Switching Centre). —Preceding unsigned comment added by 82.40.157.27 (talk) 11:53, 24 January 2008 (UTC)

Goodness, for a moment this evening I suspected User:Guy Harris was following me around and finishing all my unfinished edits, but no, this was just my self flattery. Not a problem anyway; pleasant to sit back and see competent hands do the work. Anyway, I would hate to trim the large and antiquated North American hierarchy, and indeed Guy's proposed Via net loss article merger threatens to expanded it further. Perhaps this large section would loom less "significant" or prominent if it were placed after the much smaller European sections. Jim.henderson (talk) 04:11, 3 February 2008 (UTC)


 * About the only competence I can claim is recognizing that this article and Via Net Loss both give the 5 levels of the switching hierarchy. :-)


 * Perhaps there should be a separate article (or articles) for the architecture of the North American switch hierarchy, and either give a summary here, pointing to the main article, along with the sections on the UK and French hierarchies, or split those into articles of their own as well? Guy Harris (talk) 04:27, 3 February 2008 (UTC)


 * It used to be separate, poorly titled as Office classification. My fellow grey-bearded telephone veterans in 2006 had described the hierarchy similarly in several articles, and then along came me, less competent in topic than in editing.  To a merger proposal I first said no, then was persuaded and the merger at first justified my fears and then the merged article was greatly improved during 2007.  Lately it has acquired much new and mostly good material, unbalancing the article and raising the question of reversing the merger.
 * So, yes, a split might work well, but first someone must either invent a better name, or find an existing article that would be be improved by moving either all three national obsolete hierarchies into it, and hoping for more countries, or just moving the American half of this one there. Well, I don't much like this last idea.  These national routing systems were perhaps not the soul of the PSTN in their day, and are very much not so today, and they were pretty much the same with interesting variations.  Jim.henderson (talk) 07:36, 6 February 2008 (UTC)


 * Historic Switching Hierarchies? I was largely responsible for the increase in the historic US/Canada material, and for the UK and French additions. In part, I stopped because most of the rest of the then-developed world was more of the same: mostly two or three levels, often with a final route through the capital city. There are a few quirks, mostly for political reasons: Germany, with some traffic handled through "the other" side or various military units, because of the geography and borders; Northern Ireland, mostly handled by BT in those days, but with some connections that sneaked through Eire switches due to geography; Southern Africa (to the extent that there was much network structure beyond South Africa), etc. But almost all of that might be even more plodding in a litany of recitation of obsolete material. The only area which was built out as extensively with circuit switching as North America was the small portion of Iran that got installed by AT&T before the revolution - the Shah had bought state of the art switches for the entire country, including remote villages, which required an extensive network to be designed, even though few calls would ever be presented through some of those locations. But that four-level design was never really implemented, and much of the equipment never shipped. —Preceding unsigned comment added by 76.241.139.254 (talk) 06:17, 19 February 2008 (UTC)

French Telephone Switch Hierarchy
This entire section of the article is not written in a particularly encyclopedic tone. Lordhatrus (talk) 14:10, 4 July 2008 (UTC)
 * I agree. In several places the direct or implied assertions presuppose a causality that lacks verifiable source citations, or appears to lack neutrality in that it is written as a narrative rather than a purely descriptive account.  I added a  template to the section.  Unfortunately, I lack the verifiable sources necessary to clean up the section.  Perhaps the template can cue someone else can help further. — Dgtsyb (talk) 17:33, 4 July 2008 (UTC)

Bitten the bullet
OK - I've bitten the bullet and re-organised the material. The article was dominated by a description of the hierachies which swamped the other material. I've now put that in another article called PSTN network topology. It still needs boiling down over there but at least it makes sense in its own page.

I believe now we've got a cleaner base to build out more details of the PSTN, perhaps to include things like interconnect, different carrier/operator types, relationship to the internet, standards used in its construction, political and social issues pertaining to the PSTN etc. ChrisUK (talk) 17:04, 22 March 2010 (UTC)

The article states 8 bit PCM is used.
The audio codec used on the north american PSTN is ulaw or G.711u not linear PCM. 66.114.93.6 (talk) 12:17, 24 July 2011 (UTC)


 * I fixed that. Although the definition in Pulse code modulation doesn't appear to leave much room for it, G.711 is a type of or variation of PCM. --Kvng (talk) 13:48, 26 July 2011 (UTC)


 * G.711 is a companding codec. On the PSTN, sample resolution is initially (or is sampled at) 12 or 13 bits linear PCM, then companded to fit into 8 bits.  It can be thought of a mild form of lossless compression.  Once the audio is companded, it is now non-linear PCM.  (Hence the variation)  I guess whenever someone writes of thinks of PCM, it implies the more popular linear PCM, like that of a CD.   For better clarity, I like to use/see the inclusion of "non linear PCM", companding, etc. or remove the word PCM altogether.  66.114.93.6 (talk) 16:00, 4 August 2011 (UTC)


 * That's definitely a legitimate and modern way to look at it. The original implementations of these systems used an analog compressor and a linear 8-bit converter or a non-linear 8-bit converter. In this context, perhaps you'll appreciate that such an implementation looks less like a separate encoding process and more like an alternate (non-linear) PCM system. --Kvng (talk) 20:52, 13 August 2011 (UTC)


 * Think of it this way. In a line card, there is a DAC.  The target is 8 bits, but liniear PCM has too much quantization noise so they compand it.  DACs can't sample non-linearly, so the DAC is a 12 bit sampler.  Then software then compands the adudio to fit into 8 bits.  That's how G.711 works.  Now say if you're on VoIP and using compressed codec like G.729.  It's a compression of the raw 12 bit PCM stream of the DAC, not re-compressing of the G.711 as most tend to think.  66.114.93.6 (talk) 21:38, 13 August 2011 (UTC)


 * I appreciate that's how it works today. But in the early days of digital audio, there was no such thing as a 12-bit ADC - this was a means of getting the most possible out of the 8-bit ADCs. ---Kvng (talk) 01:26, 14 August 2011 (UTC)


 * I'm with Kvng -- we both agree with 66.114.93.6 that *today* it's hard to find a non-linear DAC, so most systems use a 12 bit DAC, or better, that produces linear pulse-code modulation, then that data is companded to 8 bits. However, as described in the μ-law algorithm and analog-to-digital converter, in the early days people sampled 8 bit data directly with a non-linear 8-bit DAC (perhaps built out of an analog compandor feeding a linear 8-bit DAC). I agree that today "PCM" alone almost always means linear pulse-code modulation. I hope the "nonlinear PCM" phrasing "8-bit resolution using a special type of nonlinear pulse code modulation known as G.711" makes everyone happy. --DavidCary (talk) 19:30, 16 August 2011 (UTC)


 * Good with me. --Kvng (talk) 15:31, 18 August 2011 (UTC)

Requested move

 * The following discussion is an archived discussion of a requested move. Please do not modify it. Subsequent comments should be made in a new section on the talk page. No further edits should be made to this section. 

The result of the move request was: Not moved. Jafeluv (talk) 10:43, 27 September 2011 (UTC)

Public switched telephone network -> Public Switched Telephone Network —

Per WP:CAPS and WP:TITLE: this is a proper noun referring to the single, worldwide telephone network, per the discussion under Talk:Public_switched_telephone_network. — Dgtsyb (talk) 22:05, 20 September 2011 (UTC)


 * Oppose - Not a proper name, public switched telephone network is just a concept. Stick with the MoS - no need for an exception here. Joja  lozzo  02:20, 21 September 2011 (UTC)
 * Oppose – not even close. Dicklyon (talk) 05:48, 21 September 2011 (UTC)
 * Oppose. It is a generic name, and there have always been national telephone networks, never a single, global one (albeit they were and are interconnected). Nageh (talk) 19:52, 21 September 2011 (UTC)
 * The above discussion is preserved as an archive of a requested move. Please do not modify it. Subsequent comments should be made in a new section on this talk page. No further edits should be made to this section.

I changed caps in the lead to match caps in the title. ~KvnG 23:34, 7 June 2014 (UTC)

Proposed merge of Routing in the PSTN into Public switched telephone network
Routing in the PSTN doesn't contain enough information to warrant having its own page, but it's definitely important enough to warrant merging back into Public switched telephone network as a new section. Nathan2055talk - contribs 00:17, 5 July 2020 (UTC)


 * Routing in the PSTN is 8,355 bytes and Public switched telephone network is 11,055 bytes. The proposal paints a different picture. ~Kvng (talk) 19:21, 6 May 2021 (UTC)
 * Support merge per nom, given that the sum of the page lengths (without even accounting for the readable prose correction, and 11k+8k = approx 20k) is much less than 40k, so length alone does not justify division. Klbrain (talk) 12:15, 9 July 2021 (UTC)
 * I'm concerned about a potential WP:UNDUE issue here since Routing in the PSTN and Public switched telephone network have similar sizes. I'm not concerned enough to weigh in with an objection but concerned enough that I'd like to hear how this could be addressed before weighing in with support. ~Kvng (talk) 13:16, 12 July 2021 (UTC)
 * I've deleted the contested section (apart from one referenced section) which now means that Routing in the PSTN is of a more digestible size and should fit neatly into Public switched telephone network, as a separate section above the Technology section (I suggest). Klbrain (talk) 14:47, 10 August 2021 (UTC)
 * given that you've reversed the deletion (for reasons I understand), I wonder if you could comment here on the merge. Also, isn't it better to just delete that section and link to routing where such general background is better developed (rather than maintain/improve a poor duplicate)? Klbrain (talk) 13:51, 18 August 2021 (UTC)
 * Reverted to the 10th August version; I maintain support for the merge. Klbrain (talk) 13:56, 25 September 2021 (UTC)


 * Oppose for now. Both articles need work but I don't see that combining them now would help move that work along. Routing in the PSTN does contain enough information to warrant having its own page. Just because the merge doesn't produce an oversized article is not a good reason to do it. Routing in cellular networks needs consideration in this proposal. ~Kvng (talk) 02:52, 30 September 2021 (UTC)

Is the PSTN really just limited to the circuit-switched telephony operators?
Should the opening sentence remove "circuit-switched" and instead simply say "... world's switched telephone networks..." to more accurately reflect the current and future state of the PSTN?

"The public switched telephone network (PSTN) is the aggregate of the world's switched telephone networks that are operated by national, regional, or local telephony operators, providing infrastructure and services for public telecommunication."

If not, then the following sentence should be modified to provide the name of whatever the global telephone network is now called since many operators would no longer be a part of the PSTN since they are moving from circuit-switched to packet-switched telephony in both their core network and access lines:

"Originally a network of fixed-line analog telephone systems, the PSTN is now almost entirely digital in its core network and includes mobile[1] and other networks, as well as fixed telephones.[1]"

Isthismiclive (talk) 07:17, 9 August 2020 (UTC)


 * The manner to resolve this is not by more opinion but by providing reliable references that explain the state of the matter and contemporary understanding. The existing article is certainly already weak in such sources, and a more authoritative foundation is desirable. Kbrose (talk) 16:19, 9 August 2020 (UTC)

Come for us celine wambui and claire njeri in kenya — Preceding unsigned comment added by 105.160.107.6 (talk) 15:12, 7 July 2022 (UTC)

Retirement of PSTN/POTS
Many countries are currently in the process of winding-down/retiring their analogue last mile PSTN/POTS services, including New Zealand, UK, Iceland, Belgium and more. They are replaced by VoIP services.

Maybe there is a case to open a section on this page or even a new page for the 'Retirement of PSTN by country', similar to 'Digital television transition' Drumstick21 (talk) 18:24, 24 May 2023 (UTC)


 * Use your research and writing skills. Retirement of POTS is certainly an aspect of the modern PSTN, really the last remaining segment of the old PSTN. Long-haul routes have long been VoIP-provisioned or have used other fiber-based technologies. kbrose (talk) 18:33, 24 May 2023 (UTC)


 * I've started an article at United Kingdom PSTN switch-off. Are any other countries doing/have done similar things? &mdash; The Anome (talk) 15:35, 4 October 2023 (UTC)

International PSTN interconnect: how?
Speaking of the global PSTN, I'd be fascinated to know how VoIP-based telcos interconnect with other countries which still have circuit-switched telecoms systems -- if there are any, that is; are international circuit-switched trunks still a thing, or is international interconnection all VoIP now?

I'm thinking of places like Bhutan, Chad, etc, which are generally late on the uptake of new technologies. Do they have large-scale VoIP gateways for this? Would it be, for example, cheaper for a UK telco to haul VoIP traffic to a nearny international PoP and present it to the local telco via a VoIP gateway there? Or do they do TDM-over-IP to these locations and have a UK-based VoIP gateway, and just a couple of routers and plug-in 1U Ethernet IP-to-TDM boxes at the far end? &mdash; The Anome (talk) 15:34, 4 October 2023 (UTC)

"almost entirely digital in its core network"
The article uses the phrase "almost entirely digital in its core network". Is anyone aware of any copper long-distance trunks still existing, anywhere in the world, even in countries such as Somalia? Forget technology transition; copper thieves will have done the job by now, and it's an absolute no-brainer to put cheap and worthless-if-stolen fibre in its place, even if it's just something like an E1. 2A00:23C8:72A8:E001:DF46:CAB9:102C:4B0 (talk) 15:46, 14 October 2023 (UTC)