An energy conservation optimization strategy for wireless sensor network node based on Q-learning

Energy conservation for a sensor node is one of the key challenges in wireless sensor networks (WSN). In this paper the optimal energy optimization problem for the sensor node is concerned, and the objective is to obtain the long-term average maximum throughput per energy consumption. First, we analyze the energy conservation optimization problem in the background of cross-layer adaptive transmission over fading channels. For less energy consumption on data sensing, receiving and sending we refer to a mechanism which dynamically turns on or off different components, adjusts transmit power and modulation level of the sensor node while maintaining required performance. Then, the problem is modified by introducing dynamic power management (DPM) technique and modeled as an average reward Markov decision process (AR-MDP). Combined with simulated annealing (SA), Q learning algorithm is proposed to solve the energy conservation optimization problem with average performance criteria. Finally, the simulation results show that the approach in this paper is more efficient than always on policy or random policy. As the approach the energy consumption can be well balanced while the throughput of the sensor node doesn't decrease significantly.