Wednesday, January 26, 2005

Is the number up for the PSTN?

This article appeared in Networked Comms Insight during 2003

The title of the article asks two questions, depending which way you read it. When will the PSTN disappear, to be replaced by something else, which for argument’s sake, I’ll call the PNCN? Or, when will we stop having to use hard to remember numbers, to communicate with others?

I’d like to think the two go hand in hand, but I’ll cover off the first one first. The PNCN is already here- the Internet is the Public Non-switched Communication Network. The four letters warrant closer examination as each has an interesting parallel with the past.

Starting with “Network”, this simply implies devices connected together. In the case of a private network, it is a firmly closed user group, a public one accessible to all (based on availability of connection and ability to pay). The PSTN is global in reach but the only common denominator is 300Hz-3.4kHz analogue speech for basic telephony, everything else we take for granted in the first world (such as 64k clear channel) patchy at best in the third world. By comparison, the Internet is also Global in reach but the throughput and bandwidth available is rather more difficult to predict. We have also covered off “Public” by default here but we can take it as read that despotic regimes aside, where people want to get access they will be able to do so, mostly unfettered by the regulation that held back Telecoms post-nationalisation.

“Non-switched” vs “switched” – is of negligible interest to most users. When you talk to someone, you don’t care if your speech is packetised, encrypted, sent by multiple methods or indeed played to dolphins at sea World through underwater ultrasonic speakers en-route provided that it is something resembling commercial speech to both parties. However to the carrier, it is much more problematic. With the PSTN, the engineering comes in the unswitched capacity, once you managed to get through, you generally kept the connection as long as you needed it and the quality remained the same provided wireless wasn’t involved. With the PNCN, however, as it currently involves multiple technologies, effective capacity management must be an operational nightmare.

Finally, we come to “Telephone” versus “Communication”. The PSTN is admirably suited to talking and can be used for low speed data, whether bonded multiple 64k channels, or 300 baud modems. The Internet as a whole, however, is pretty hopeless at coping with speech from end to end unless you are prepared to accept random quality and poor reliability levels. As for real-time video, forget it.

That is not to say that significant chunks of the Internet cannot be used for carrier class telephony within the same supplier. Indeed, how many Carriers would swear on the bible that all of their private circuits provided to Customers as such are actually genuine, dedicated, uncontended provision?

Now, onto the thorny subject of numbers. We take telephone numbers for granted and they are comparatively easy for the public to understand and use. If they are so simple, why don’t we use them on emails? Well, in the early days of Compuserve, that was the case. If you wanted to send me a message, I was something like [123.456789] (I didn’t actually want a Compuserve account, having jumped into the Internet deep end with Demon when the setup was fairly complex, but it came free with a project I was working on). Account names often had semi-random alpha numerics in them, generally at the whim of a Sysadmin because simplicity and scalability hadn’t really been thought through at the time. These days, the de-facto standard seems to be firstname.lastname@company.type which is fine if you aren’t called John Smith, or in the Gulf Countries, Mohammed Abdullah. So why don’t we do the same with telephony?

Well, the reality is that we do. In our diaries, on our mobiles, on our featurephones, we create an alias that we immediately recognise and can associate with who we want to contact. We can remember a handful of numbers but there comes a point where we have to look them up, whether on a piece of paper, or having to dial 118£££. With instant messaging, we create “buddies”. With email, we create distribution lists or mailing lists. With favourite websites, we create shortcuts. However, it is not normal to number them!

I bought a couple of cheap phones on ebay recently to replace my home caller display units that were getting unreliable. After living with them for a while, something struck me about how the speed dialling worked that was unusual for a wireline instrument. Rather than bring up a numbered list, the list was sorted alphabetically. It did mean that you couldn’t do < 3> <2> for entry 32, but instead, when you pressed speed dial, you simply chose who you wanted in the same way that you type SMS on most phones, i.e. to access entries beginning with K, you pressed 5 twice. The phone positively encouraged you to label and store calls, both incoming and outgoing. Whilst this was fine, I eventually realised that I had to do exactly the same chores again on the upstairs set!

So, are numbers better than meaningful names? Yes and no. Numbers are quicker, but harder to remember. Text is more straight-forward, but requires greater accuracy. Dyslexia affects numbers as well as names, we get a lot of calls to a test phone on 01422 xxxxxx when the caller meant to dial 01442 xxxxxx, i.e. Hemel Hempstead, not Halifax!

ENUM is an interesting development, which is an attempt to provide a mechanism to contact someone by any channel. It works a bit like the Internet DNS lookup that turns a URL such as into an IP address such as In simple terms, you ring, fax, email or message the one number, which is then translated to the appropriate address for delivery.

I see the future as including both- When you want to contact someone from a PNCN device, you find who you want quickly and intuitively from contextual menus guided by what information you already know. However, if you are in a Bedouin village in the Sahara, you find the ENUM from the community solar powered PC like-device, then queue up for your turn on the rotary telephone. (Or buy a journalist a large drink and borrow his Satcom system!)

So is the number up for the PSTN? Not yet. Too many security problems on the emerging networks at the moment. Too many technical issues, Telcos feeling the pinch, not enough investment. But, why should there only be one internet?

No comments: