06 July 2014

Softwarization paving the way to Cognition-as-a-Service

In Computer Science abstractions hide hardware specific details which are considered not useful at a given level. The OS in fact stands between the Users’ programs and the physical computer hardware, it provides  high level functions (e.g., allocating memory or writing to disk). In this sense, it has been mentioned that SDN can be see seen as a Network Operating System (even if there are interpretation of SDN as a Compiler).

Let’s imagine now to extend this concept (see my previous post) to a distributed “5G Operating System“ capable of spanning from the Terminals, to the Network to the Cloud. Not only that: imagine even embedding cognitive capabilities in the 5G OS, so that cognitive apps can be easily developed by means of APIs, rather than requiring Developers to reinvent the wheel any-time.

Welcome to the world of “Cognition as a Service” (CaaS). By the way, isn’t this also the direction of the next generation cognitive management systems (in the sense of OSS/BSS) that we’re looking for ?

In fact, it will not be possible anymore adopting traditional management approaches (with declared objectives and observed behaviour) for 5G. Complexity and dynamism will be too high. Dynamic or static modeling for (open or closed loop) control will become very complicated and unstable if not supplemented with a variety of methods and control techniques, including (nonlinear) dynamic systems, computational intelligence, intelligent control (adaptive control, learning models, neural networks, fuzzy systems, evolutionary and genetic algorithms), and artificial intelligence.

This trend is already detectable today. It’s not surprising, for example, that a companies like OTTs are pushing the idea of cloud robotics. It is no more Science Fiction when we say that we’ll be living soon with cognitive machines and robots.