An Energy-efficiency Node Scheduling Game Based on Task Prediction in WSNs

For wireless sensor networks, unbalanced task load will decrease the lifetime of network. In this paper, we investigate how to schedule the sensor nodes to sleep or wakeup according to the dynamically changing task load. We first demonstrate that for a sensor network with uniform node distribution and constant data reporting, balancing the task load of the whole network cannot be realized. Then we define the concept of state transition and design a state transition model for sensor nodes. By introducing Markov chain, we further propose a task prediction method to predict the local task load in the next time period. Finally, we propose an energy-efficiency node scheduling algorithm based on game theory (ENSG) for WSNs. To obtain better performance, the residual energy of sensor nodes and local task load are both considered into the payoff function of our game. Our simulation results show that ENSG can guarantee the real-time task completion and prolong the lifetime of network.