Research On Target Tracking And System In Wireless Sensor Networks Based On Character Of Movement And Terrain Restriction

Sensor Networks is composed of some sensors in which communication chip is embedded. Sensors are deployed in a needed place .After the deployment , sensors will collect the physical data in the networks'environment. Then send the data they have collected to the users of network. Sensors in Ad-hoc style may corporate to provide some complex functions for the users.Using large numbers of tiny sensor nodes which are organized in Ad-hoc style, wireless sensor networks can track target in battlefield rapidly, precisely and snugly. Consequently, how to track military targets in sensor networks is becoming hotspot in research. The dissertation develops research deeply about how to track target effectively and how to impove energy efficiency in process of tracking using trajectory predication.The dissertation founds the prediction model for direction of target movement with movement model of targets. Based on the hypothesis that target moves in line in local area, self-adaptive least square method is adopted to to predict the direction and speed of target. The dissertation uses the method of data fusion to classify the targets. Bring on the classification algorithm which is proved to be right with the experiment.In allusion to terrain's effect on target movement, the concept of Terrain Accommodative Probability of target is introduced and a markov model for updating Terrain Accommodative Probability. of target is founded. In this way, we can update posterior Terrain Accommodative Probability of target online and according revise the prediction result of the model based on target movement. On the base of the model of the target classification, the dissertation research on algorithms available for target location and tracking. To improve the precision and reduce the complicacy and the communication spending, design locating algorithm based on the classification, with nodes location force to the target location. With the experiment results, this algorithm has more precise than the normal one, and less requirement for the hardware. Focused on the first problem of WSN, the dissertation purpose a tracking algorithm based on target classification model. By awaking less nodes the energy of targets is effectively saved and the life time of networks is prolonged.In the end,the dissertation expatiate the design and implementation of prototype system based on Mica2 platform to verify key technologies such as target detecing and localization algorithm, and the design test case of the system.