ALOHA and Q-Learning based medium access control for Wireless Sensor Networks

Abstract

This paper presents a novel Medium Access Control (MAC) protocol named ALOHA and Q-Learning based MAC with Informed Receiving (ALOHA-QIR) for Wireless Sensor Networks (WSNs). This scheme is implemented by combining frame based Slotted ALOHA and Q-Learning. ALOHA based schemes have the benefits of simplicity, low computation and overheads, but suffer collisions from the blind transmission strategy. By considering the previous transmission experience, Q-Learning offers nodes certain intelligence to access slots with a lower probability of collision. During data transmissions, receivers are informed by transmitters about their future slot selections so that the receivers can turn off their radio in other slots to save energy. Compared with basic Slotted ALOHA, ALOHA-QIR obtains significant improvements in throughput, delay and energy efficiency.