Optimal policies based on QoS for adaptive communication system with Markov Decision Process

Modern communications, especially mobile communications, face unpredictable environments. Adaptation to unpredictable environments is necessary for modern communications. The communication systems self-tuning to track changes in the environments are called the adaptive communication systems. The systems can be viewed as a special kind of feedback control systems with the elements of sensor, decision-element or controller, and actuator. This paper focuses on the decision-element. The decision-element uses optimal policies based on QoS with Markov decision process, which seek to satisfy a set of system QoS. The simulation results show that communication system adaptation can be guided by decision rules based on QoS found by the methods of Markov decision processes, when there are multiple controlled features, and multiple system measures used to drive the decisions.