A tutorial on reinforcement learning in selected aspects of communications and networking

Telecommunication systems are increasingly complex, dynamic, and heterogeneous. Tools are needed to efficiently support and automate complex control and management processes. Reinforcement learning becomes one of the most attractive and popular solutions applicable to a wide variety of different aspects of communications and networking. Its development is further boosted by the favorable conditions created by both the existing IT infrastructure and evolving network architectures.