Towards Cognitive Autonomous Networks

11System Aspects for Cognitive Autonomous Networks

Stephen S. Mwanje¹, Janne Ali‐Tolppa¹, and Ilaria Malanchini²

¹Nokia Bell Labs, Munich, Germany

²Nokia Bell Labs, Stuttgart, Germany

Network Management (NM) is a system of devices and software functions with the respective interfaces to allow for monitoring, control, and configuration of network devices. Network Management Automation (NMA) extends this system by adding automation software functions both within the network devices and in the network control devices. Beyond the individual functionalities for configuration, optimization, and healing (as discussed in Chapters 7–10), there are system‐wide challenges that also need to be addressed. The most widely discussed challenge here is self‐organizing networks (SON) coordination for which early solutions focused on prioritization of SON functions (SFs) at run‐time. Cognitive Network Management (CNM) takes the perspective that such system‐wide challenges can be addressed using cognition as the basis for decision making.

This chapter discusses these system‐wide challenges in the CNM environment where cognition is the basis for functional development. It seeks to address two core questions: (i) what are the core ideas that can be added to the SON environment to advance it towards more cognition and autonomy? (ii) What are the system‐wide challenges that need to be addressed when the functions are themselves cognitive and how should such a cognition system be managed or leveraged? ...