Research On Target Tracking Algorithm In Distributed Wireless Sensor Networks

Wireless sensor network (WSN), which integrates the technology of modern sensor,wireless communications, microelectronics, embedded computing and distributedinformation processing, is a new interdisciplinary of building universal newenvironment and realizing universal control. WSN is suitable apply to target trackingsystem, which has the feature of low power consumption, low cost, concealment androbustness. Target tracking technology has been widely applied to national security,traffic management and other disaster areas, so low energy consumption and highreliability target tracking algorithm has theoretical value and practical value.The distributed target tracking for research object is studied, which is a keytechnology in WSN. An improved algorithm is proposed in this paper, which can solvethe question of unbalanced communication, low tracking quality and high algorithmcomplexity. The main points are as follows:The existing sleep scheduling protocol can’t meet the needs of target tracking.Therefore, a sleep schedule protocol for target tracking in sensor network (TTSSP) isproposed in this paper. Through the establishment of sensor network model and nodeperception signal model, TTSSP divided different sensor node’s work mode in eachtracking stages and introduced in detail the node before target emergence and the nodetarget tracking and how to wake up the sensor node when the mobile target lost. Thesimulation results show that the algorithm can reduce the energy consumption, improvethe successful tracking rate and ensure the tracking quality.TTSSP target tracking algorithm is proposed in this paper, which based on theTTSSP protocol. The weighted centroid localization algorithm and target locationprediction are used to realize target location, which based on distributed static clusternetwork structure. The target forecast divided prediction in the whole network into nodein cluster and sensor between clusters. The cluster head node can predict next timetarget’s location and other related information according to the known locations andwake up the sensor nodes in that area. Before the time target run in the current clusterexhaust, the neighbor cluster head is awaken according to forecasting information. Thesimulation results show that the algorithm with merits of high precision, low targetmissing rate and low energy consumption.