Reinforcement learning based energy efficient protocol for wireless multimedia sensor networks

With the advancements in sensor networks, Wireless multimedia sensor networks (WMSNs) have emerged and shifted the objectives of sensor nodes to multimedia devices which can retrieve audio, images, and video. In WMSNs, the sensor nodes are tiny microphones and cameras which can transmit image, audio or video using the network. However, these nodes are battery constrained (i.e., may become dead after passing certain iterations). Therefore, improvement of the network lifetime is a challenging issue of WMSNs. In this paper, a reinforcement-based energy-aware protocol is designed and implemented. To successfully implement the reinforcement-based protocol, a State-Action-Reward-State-Action (SARSA) is used for learning a Markov decision process. Extensive experiments are considered to evaluate the significant improvement of the proposed protocol. Comparisons are also drawn between the competitive protocols and the proposed protocol. From comparative analysis, it is found that the proposed protocol conserves more energy as compared to the competitive protocols.