Internet is likely to have a great impact on our future daily lives progressively becoming inherent part of strategic areas such as communications, electricity control, transportation, utilities and resources management, healthcare, manufacturing, commerce, etc. In the future, any smart object with embedded communication capabilities will be able to receive and transmit information to and from other devices (of humans and machines) and objects through omni-pervasive communication networks. And this, maybe, will enable what Clay Shirky calls (in his the new book) “Cognitive Surplus”. But, my question is will/could these networks have some cognitive features inherently hardwired?
The first definition of a cognitive network (rather than from the cognitive radio community) comes perhaps from D. Clark  “a network that can assemble itself given high level instructions, reassemble itself as requirements change, automatically discover when something goes wrong, and automatically fix a detected problem or explain why it cannot do so”.
In  a more recent definition is given: “a cognitive network is a network with a cognitive process that can perceive current network conditions, and then plan, decide and act on those conditions. The network can learn from these adaptations and use them to make future decisions, all while taking into account end-to-end goals.”
Looking at these definitions, one might expect that to make a network cognitive it is necessary to exploit a knowledge plane or some cognitive processes (for instance, based on reasoning, learning, etc). Well, a quite large international community is working by adopting this perspective. As usual, there are still the dilemma questions on costs/benefits: what is the “price of the feature”? What are the (capex, opex) savings?
I’m organizing a networking session at ICT2010 about this.
I’ll tell you about the results.
On the other hand, it sounds also reasonable to observe that the increasing complexity of future networks (as a consequence of pervasive connectivity, diffusion of more and more powerful mobile devices, smart objects/things interconnections, etc) will results in a cognitive-like behaviour, almost spontaneously.
It’s rather easy to understand that the “networks within networks” in human brain are the results of a long evolution. It is also well known that adaptation is the heart of this evolution: a brain, coming from millions of years of evolution, is able to perform complex cognitive processes assuring human adaptability and survivability. Whatever will be the level of abstraction one may want to consider to model the brain (e.g., multi-level networks of molecules, organelle, neurons…), in any case, there are some basic hard-wired rules embedded into the interconnection capability of the said abstraction elements (at the different levels). This is observed also in most animals: they seem to have a set of rules that they follow under certain circumstances, sort of a hard-wired stimulus/response primitives. Then, as a result of interactions between group members, (for example ants) the community as a whole eventually will exhibit patterns called “emergent behaviours”. In  some examples of bio-inspired primitives are reported: integrity and variability, positive feedback, negative feedback, response thresholds, leadership, redundancy, synchronisation and selﬁshness. This is just an example. For humans, for example, this set of hard-wired behaviors is very small, and generally plays little role in day-to-day existence. But now we’re talking about network enabling human collective intelligence and new related ecosystems.
So, my point is: which is the way to look at those “rules” to be embedded into nodes, devices and objects of future networks (maybe at the different interacting levels) in order to allow a spontaneous emergence of (certain desired) cognitive-like behaviours?
By the way, this approach prefers considering cognition as a side-effect emergence in future networks, rather than adopting a traditional top-down formal design.
 D. D. Clark, C. Partridge, J. C. Ramming, and J. T. Wroclawski, “A knowledge plane for the Internet,” in Proc. of SIGCOMM ’03;
 R. W. Thomas, L. A. DaSilva, and A. B. Mackenzie, “Cognitive networks,” in Proc. of IEEE DySPAN 2005, pp. 352–360, Nov. 2005;
 D. J. T. Sumpter. The principles of collective animal behaviour. Philosophical transactions of the Royal Society of London, 361:5–22, 2006.