Energy-Efficient Distributed Multi-Sensor Scheduling Based on Energy Balance in Wireless Sensor Networks

An energy-efficient distributed multi-sensor scheduling (EDMS) scheme is proposed for target tracking in wireless sensor networks. To increase tracking accuracy, a distributed unscented Kalman filter (UKF) is developed to estimate the trajectory of the target and predict error covariance matrix for next time step. To maximize the lifetime of the whole network, a novel energy balance model is constructed. In the EDMS scheme, multiple sensors are dynamically scheduled as tasking nodes to form a dynamic cluster based on the minimal trace of error covariance matrix and a cluster head is selected based on the energy balance model. In addition, adaptive sampling intervals are adopted to save energy consumption under the satisfactory tracking accuracy. The simulation results indicate that the tracking accuracy of the proposed EDMS using distributed UKF is significantly improved compared with extended Kalman filter (EKF). Under the satisfactory tracking accuracy, the EDMS scheme considering energy balance has better energy balance distribution than that not considering energy balance.