Research On Target Tracking Protocol In WSN

Wireless Sensor Network (WSN) is an emerging network with advantages of self-organizing,dynamic networking and high fault-tolerance etc, and it has applications of extremely broad field. Inmany applications of WSN, moving target tracking is a very important one, and the research on targettracking technology has become one of the current hot spots in WSN.Dynamic tracking group is a commonly-used tracking mode in target tracking application,however, the energy consumption of cluster head (CH) is large in this method, and the failure of CHwill easily lead to the loss of target. To solve the problem above, the dynamic cluster structure hasbeen improved and a new type node called auxiliary node (AN) is designed in this paper. The ANloads the main computation tasks of the traditional CH to reduce the energy consumption of it, in thisway, the loads of cluster become more balanced. On this basis, a two-node collaborative trackingprotocol (TNCT) is proposed, and TNCT elaborates on the building process of the improved cluster,provides the specific rules for selecting the CH and AN, redefines the responsibilities of cluster nodes,and achieves the two-node collaborative target tracking algorithm through the cooperation betweenCH and AN. To improve the tracking accuracy, the Extended Kalman Filter (EKF) algorithm isexecuted in CH to evaluate and predict the target trajectory.With respect to the velocity changing issue, an adaptive sampling interval algorithm is given, theidea is to dynamically adjust the sampling interval to track based on the current target velocity,combined with the variable speed range of the target. Decreasing the sampling interval to reduce thetarget loss probability when the target is in acceleration, while increasing the sampling interval toreduce the energy consumption when the target is in deceleration or stationary, and this algorithm canenhance the adaptivity and energy-efficiency of tracking. To further reduce the adverse effects causedby the failure of CH and AN, a lightweight fault-tolerant mechanism for tracking group is proposed inthis paper, and this mechanism can achieve the fast transfer of CH while the CH is dead, and canreduce the target loss by supporting the fault-tolerance for CH and AN of the cluster.Eventually, OMNET++platform is applied to simulate the TNCT protocol. Simulation resultsshow that on the premise of higher tracking accuracy, the energy consumption of TNCT protocol ismore balanced and the target loss probability has been greatly decreased.