On the implementation of Cognitive Autonomous Networks

Cognitive Autonomous Networks (CAN) is a promising solution for next generation network management automation and it replaces state-of-the-art Self Organizing Networks (SON) quite successfully. In CAN, a set of Cognitive Functions (CFs), which replace the existing SON Functions (SFs), automate the network processes under supervision of a controller. The CFs interact with the environment to learn and decide suitable network configurations to optimize their objectives, which they send back to the Controller. The Controller works on these propositions and calculates the final value which is of the best interest for all the CFs. Learning capabilities of the CFs help CAN overcome the problems faced by SON, but, at the same time, makes the implementation of CAN very challenging. In this Use Case Letter we elaborate how CAN can be implemented for experimental purposes.