The rise of the stupid networks 2.0

I recently re-read David Isenberg‘s seminal paper The rise of the stupid networks which explains the End to End principle . The paper applied originally to IN (Intelligent networks) for voice but today as we shall see below, the same principles apply to IMS for Mobile data applications . This is sad – and explains the low uptake of IMS applications. The only solution is to treat the network as a black box.

The Rise of the stupid networks – a brief synopsis

Rise of the Stupid Network by David Isenberg is a seminal paper which describes the End to End principle. As you read the synopsis below from that paper, remind yourselves that it applies to IN (and note the similarities between IMS).

‘Intelligent’ networks arose with the idea of doing digital switching for voice calls. Driven by the motivation of ‘vendor independence’ and the prospect of getting better deals from their suppliers by fostering interoperability between suppliers; Intelligent networks gained momentum.

In contrast to interoperability between the network and the supplier, interoperability between the network and the customer was managed by limited, carefully designed interfaces. The Intelligence of the network itself was limited to the intelligence of the designers – and hence was confined to what the designers knew best – voice call routing, calling party connect – that sort of thing

The impact of not talking to their customers meant that new customer needs filtered only slowly into the network. More importantly – they needed changes to the network. Consequently, so as to minimize changes to the network, only the needs with the biggest payoff get attention – a risk free, innovation averse strategy. This leads to business cases, ROI, development plans etc etc. This can take years.

The Edge of the network

In contrast, any innovation that resides at the edge of the network (and treats the network as a black box) does not suffer from the same limitations. The success of ‘edge of network’ applications does not demand any changes to the network and does not depend on the evolution of the network.

So, what do did we see? Because the real customers cannot drive innovation and ensure that their needs are being heard (think limited, controlled interfaces as before) .. vendors drive the hype in the guise of customers. This makes matters worse since there is a perception of innovative services .. none of which the real customer wants.

Driving innovation from within the network leads to an odd situation of ideas which depend on the network to catch up. This is never a good place for innovators to be since there is no ‘one’ network(and even within a country there are often many networks). Evolution of networks is unpredictable. It takes too long. It is too risky to tie your business plan to a specific future direction a network may take.

Wearable computing

For instance – At the mobile world congress in Barcelona 2008 – one operator mentioned that ‘wearable devices’ was a key trend to watch

This is perplexing .. Wearable computing? Are we honestly thinking of wearable computing as a major trend for devices? Should developers start to plan for it? Should they invest their time(and scare resources) on wearable computing applications?

Are there startups in a garage now plotting on the next killer ‘wearable device’ application?

Perhaps I am missing something .. but I doubt it ..

The point is this: Do we expect the entire network to evolve in this direction? Maybe a specific operator may announce an initiative – but will that matter to innovation in general? Will it lead to critical mass? What about interoperability?

This is an example of innovation driven from within the network – and even if there is a need for wearable computing – will it make commercial sense for innovators to focus on it (considering the issues of interoperability and critical mass).

Sadly, this is IN 2.0. It does not matter if its voice or data – the problem remains two fold

a) Building intelligence in the network limits the innovation to the imagination of the network provider who view each problem within the context of their existing capabilities and

b) The mindset of Intelligent networks keeps the customer driven innovation out of the network

This leads to a negative spiral – which we saw with IN, IMS and other esoteric propositions like wearable devices alongwih any terms like 3G applications, wireless applications, wimax applications, LTE applications and so on (i.e. by definition they tie intelligence and services to a specific network type)


The only solution is to treat the network as a black box and focus on edge of the network innovation. This is also likely to mean that innovation will arise from players outside the current ecosystem – hence look to Apple, Google and others to create network agnostic innovation.


  1. Dean Procter says:

    He’s on the money there. Exactly my strategy. We’re doing it with the mobile and financial networks right now closely followed by the broadcast networks.

  2. Ajit Jaokar says:

    many thanks Dean! Keep us posted about your work kind rgds Ajit