A Reinforcement Learning Admission Control for Wireless Next Generation Networks

In this paper we study the call admission control problem to optimize the network providers revenue guaranteeing the quality of service to the end users. We consider a network scenario where each class of service is characterized by different parameters and an associated static revenue. We formulate the problem as a Semi-Markov Decision Process, and we use a real time Reinforcement Learning algorithm. Other traditional algorithms require explicit state transition models, instead RL learns model of environment from experience. We show that RL policy provides better solution than classic policy as the greedy algorithm