Skip to Main Content
This paper presents a novel routing protocol based on the Learning Automata method for large scale Wireless Sensor Networks (WSNs) codenamed DRLR (distributed reinforcement learning routing). In this method, each node is equipped with learning automata so that it can learn the best path to transmit data toward the sink. The approach proved to be efficient, reliable, and scalable. It also prevents routing hole by considering network density and average of energy levels available. The approach also increases network lifetime by balancing energy consumption. We compared our approach to two other methods (MMSPEED and EESPEED) and the simulation results show our algorithm to better meet end-to-end delay and reliability requirements and to improve network lifetime more.