28 August 2015

Nature says that three control loops should be enough

Future Telecommunications infrastructures will be composed by such a large number of highly dynamic nodes, systems and devices that the number of connections will require going far beyond today’s paradigms. In the long term, Softwarization will transform said infrastructures in complex systems of virtual entities, pervasively distributed in our environment.

At the edges (where start-ups are booming today) it is likely that we'll see highly dynamic games of virtual resources (belonging to the same, or different Operators or any other Players), executing any sort of services by using local-edge-centralised processing and storage capabilities.  

It is pretty clear that it will not be possible any more adopting traditional management and control. 

Modelling (open or closed loop) control will become too much complicated and unstable if not supplemented with a variety of novel control techniques, including (non-linear) dynamic systems, computational intelligence, intelligent control (adaptive control, learning models, neural networks, fuzzy systems, evolutionary and genetic algorithms), and artificial intelligence. 

Out of this “chaos” of interactions, a sort of collective intelligence will emerge. This reminds the collective intelligence emerging in a beehive. Also, this is bringing me back when I studied the theory of “enactive” behavior of living systems, developed by F. Varela.

The theory (quite influential in some Artifical Intelligence avenues) argued that adaptive behaviour of simple living entities (e.g., bees) is based on two interrelated dimensions: 1) perception consisting of perceptually guided action and 2) cognitive structures, emerging from the recurrent sensori-motor patterns, enabling action to be perceptually guided.

So simple living entities cross several and diverse cognitive domains (or micro-worlds) which are generated from their interactions with the external environment: within a micro-world the behaviour is simply determined by pre-defined sensori-motor loops, simple and fast; from time to time breakdowns occur which are unexpected disruptive situations determining the need to change from a cognitive domain (i.e., from a micro-world) to another one. Importantly, this bridging (during breakdowns) is assured by the nervous system (allowing a new adaptation and the consequent learning of new sensori-motor loops).

Amazingly, this behaviour has been successfully exploited by Nature with three nested control loops (automatic/sensori-motor, autonomic/nervous, global). I've been impressed realizing that these same principles (e.g., adaptive control based on three control loops, supported by learning) are used for developing auto-pilots, smarter robots...and self-driving cars! 

The same principles can be used for decentralising the "intelligence" required for the execution of services (and data storage) in Fog - Edge - Cloud Computing resources in a SD-I.