Years ago, before the internet became the phenomenon that it is now, David Isen put forward the idea of the Stupid Network. In a short and very readable paper he laid out the basic tenets of the concept and its benefits. Benefits included driving innovation by accelerated deployment of new technologies, and this was to be driven by unlocking control of the network. In essence, the network would continue to support foundational functions such as transport and routing, while service intelligence was to be located outside the network.
This certainly did not endear him to his employer, AT&T. The telco business model was, and is, strongly dependent on maintaining control over the network, and therefore the services it enables. They were right to be concerned. All these years later, having nearly achieved what was called the death of distance, much of the telco's profits come from features above and beyond the basic business of making voice and data connections.
Now, before we look at the competing concepts of intelligent versus stupid networks, it may be helpful to compare telephone networks and computer networks, and especially how the latter evolved over the years.
Up to about the early 1980s the principal model of computing reflected the underlying economics. Computers were large, expensive and required trained experts to install, operate and maintain them. Software was constrained by the hardware and communications channels. This necessitated thin clients and centralized intelligence. Therefore we found dumb terminals, often shared rather than one per person, low-speed and error-prone communication channels, and large time-sharing mainframe computers that supported dozens or even hundreds of users. Over time the terminals got increasingly intelligent and those central computers did get smaller and less expensive, yet this model of a dumb edge and a smart centre persisted for a long time.
In my youth, before I got involved in telecom, I worked in this business and got to know it very well. I changed my career focus as the technology evolved. When personal computers (PCs) began to make inroads I, along with rest of the industry, made the move. Rather than a small number of computer manufacturers like IBM, Honeywell, Burroughs and Amdahl, innovation and entrepreneurship flourished. Microsoft, Lotus and others appeared, and they and the hardware vendors sold direct to businesses and consumers. I still remember when co-workers felt an incredible degree of empowerment when, with a PC and spreadsheet right there on their desks, they could do stuff without pleading for someone (like me!) to write some software and get them access to valuable machine resources. Anyone with eyes could see right then how powerful a concept it was to have an intelligent "edge" on the computer network.
The virtuous spiral did its job well, creating even more opportunities for everyone. Within a surprisingly short span of time the mainframe was sidelined - not quite dead, but dying - and everyone had a PC on their desk.
It was inevitable that all these sophisticated yet isolated devices would leverage the network effect by starting to communicate with each other. It started with the simple things, like email, then suddenly the web was among us. The internet is today the ultimate stupid network, capturing very well the essence of Isen's vision.
It is no coincidence that telephone switching followed a similar path at first. Yes, it did lag some years behind computing but that was justified, at least partially, by the need to harden the technology for 24x7 service availability. Even so, by 1980 the switching equipment was almost all digital and computer-driven. But then matters languished for many years despite the introduction of some very sophisticated hardware and software inside the networks, far on the other side of the demarcation point, and out of the reach of consumers (see Parts 1 and 2). The phones looked prettier and did a few new things like speed dial, in-home wireless, speaker-phone and the like that did not require sophisticated signalling to the network. That was not and is not made available to residential customers.
Inside the network (PSTN or Public Switched Telephone Network) some changes were apparent. There were now feature servers and a closed signalling network for sole use by the telcos - SS7. I won't bore you with a lot of industry jargon, so let's just call it by its most generic title: the Intelligent Network, or simply IN.
As I mentioned in part 2, residential customers were not offered signalling that made use of the underlying network capabilities. There were limited exceptions, mainly targeted at enterprise customers, but also offered to competing carriers when they were forced to do so. (That I'll perhaps cover in future so that I can focus here on you and me, the typical residential customers.) So, while the network evolved, mainly providing operational and economic benefit to the telco, consumers saw very little in the way of substantial improvements. To them the network looks much the same now as it did three decades ago!
This may seem odd since surely there are lucrative opportunities to get more revenue by offering customers more or at least better services. But as I've mentioned before, in the non-competitive local telephone market that was in place pretty much everywhere in the world until the late 1990s they had little incentive to do so. It wasn't just their fault, it was also governments and regulators that share the blame of making them complacent. Again, that's another story and I do want to talk about the technology.
Which brings us, after a very long and hopefully not too sinuous a path, to IN. This became a force in the industry through the 1990s, which may seem surprising since the work on IN was virtually invisible to everyone outside of the telecom industry. Despite the massive effort put into IN by the telcos and the major (and minor) equipment vendors, in the end it was stillborn. It does have some successes, which are most notable for being so rare. For one, there's the Calling Name feature. Others include call screening features that go by various names, including Call Director and Call Intercept, and then there are a few more sophisticated services like Verizon's iobi.
Why did IN fail to fulfill its objectives? Why did consumers never see the rich set of features that IN promised? For this we have to go back and understand the telco motivations to build IN, some nasty technical issues, and in particular (to reprise the title of this series) their overarching need to keep control of the phone. The primary reason why they even bothered to pursue IN, and, make no mistake, it was driven by the telcos and not the equipment vendors, is probably very different from what you might guess. I'll get into that in Part 4 of this series.
Posts
No comments:
Post a Comment