Adaptive dual cluster heads collaborative target tracking in wireless sensor networks

Dynamic clustering is an effective approach in target tracking. However, the energy consumption of cluster head is relatively high, and the failure of cluster head will lead to target loss. In this paper, we proposed a new cluster with two heads which can track the targets collaboratively based on improved extended Kalman filter with a changeable sampling period. An auxiliary cluster head, selected according to the principles of maximum residual energy and shortest distance, is used to predict the target trajectory using improved EKF. The targets are tracked collaboratively by the cluster head and auxiliary head with a lightweight fault-tolerant mechanism. An adaptive sampling algorithm is designed based on average velocity estimation to change the sampling period, and the target loss probability and energy consumption are reduced significantly. The protocol is tested extensively and compared to other existing methods. The results clearly indicate the benefits of our algorithm in terms of the tracking accuracy, energy consumption and target loss probability.