Study On Wireless Sensor Networks For Field Surveillance

With the appearance and development of pervasive computing, wireless sensor networks have allured more and more eyeballs as a new type of data collecting and processing because of its wide appliance foreground. At present, many technologies of wireless sensor networks wait in an exploration stage yet, such as network coverage, transmission security, reliability, quality of service and so on. Only solving these technical questions effectively, wireless sensor networks then have the opportunity truly to apply to applications, including military affairs, medical treatment, commerce, education and environmental applications, to display the latent enormous value, and to transform truly as the productive forces. In this paper, we start from the field surveillance using wireless sensor networks, have done a few beneficial explorations and acuquired some creative research results according to its performance, strategy and methods. Our work is mainly shown in three aspects as follows:Firstly, supposing the finite ability of target detection, we propose target intruding algorithms based on local information, discard assumption for global information of sensor deployment and provide the real method for evaluating the performance of field surveillance using sensor networks. Enlightened from the traditional routing algorithm, we design and realize five path-finding algorithms and find out that FDP has the best performance through simulation. Based on the intruding algorithms of the target, we discuss the performance of field surveillance from the view of target detection probability, and point out the performance-improving direction of static sensor networks– hiding.Secondly, introducing coherent movement to wireless sensor networks, we show the optimal moving direction f_s~- and corresponding intruding direction f_t~⊥for mobile nodes and the target respectively. Moreover, we analyze the exiting Nash equilibrium between the mobile node and the target from the view of game theory. Based on the uniform moving model of the mobile node, we analysize the moving direction of mobile nodes for more degree of coverage and the intruding direction of target for less degree of exposure. Simulation result manifests the correctness of result above.Finally, putting forward the problem of interception against intruding target, we present unattended surveillance using hybrid sensor networks. We propose target intercepting method using mobile nodes with the help of static nodes sensing and computing as landmarks. Besides, according to the multiple targets - multiple mobile nodes case, we design a distributed task assignment protocol PMB based on probe message. Experimental results show that the algorithm above has good feasibility and some effectiveness.