Research On Energy-Efficient Routing Protocols And Target Tracking In Wireless Sensor Networks

In recent years, a new measurement and control network technology—wirelesssensor networks (WSN), with micro-electromechanical systems, computer science,communication, automatic control and artificial intelligence and other disciplines, hasexperienced rapid development. WSN will deploy lots of tiny sensor nodes with thecapabilities of communication and computing on the monitoring areas for real-timedetecting, acquiring and processing information of monitored objects. Through theabilities of self-organization and self-adaptation that WSN possess, and the cooperationway of multi-hop communications, WSN realize the perfect fusion of the naturalphysics information and logical world information, and change the human conventionalways of living. Because of the advantages of low power consumption, easy deployment,high fault tolerance and high precision measurement, WSN have extremely extensiveapplications in military, civil and commercial fields, and so on.According to the research of the prior work, this paper carries on research on therouting protocols, regional communications coverage of deployed sensor nodes and thetarget tracking technology in WSN from the application layer. Based on the differentapplications, some practical and valuable innovation algorithms are proposed in thefields of the load balance of networks, the optimal selection of cluster heads, thedistribution optimization of the deployed sensor nodes and movement and the preciselocation of motion objects. Then, the performances are confirmed and estimated by thecomputer simulation.The specific content is as follows:(1) According to the effect of limited radio range of sensor nodes upon routing inWSN, an improved multi-hop clustering algorithm based on minimum spanning tree(MST) is proposed to ameliorate the problems of energy consumption because of therouting. This algorithm optimizes cluster heads according to the characteristic inpoisson process of voronoi diagram, and establishes the MST as sensor nodes’ dynamicrouting for optimizing energy consumption in WSN. This algorithm establishes theMST as sensor nodes’ routing and then optimizes the distributed density of cluster headsaccording to the characteristic in poisson process of voronoi diagram. The results of thesimulation indicate that the novel algorithm is able to make energy load balanced at thepremise of cost tolerance, and prolong life-span of the networks effectively and reducethe time complexity compared with multi-hop hierarchy algorithms based onlow-energy adaptive cluster hierarchy (LEACH) during the same simulation conditions. (2) To eliminate the affect of the distribution of deployed sensor nodes on thecoverage of effective communication signals in deployed areas for wireless sensornetworks, an algorithm of Distributed Deployment of Coverage Probability (DDCP) isproposed based on the effective communication coverage in this paper. Assuring thedeployed precision, DDCP is used to not only consider the environmental influence foruncertainties of the deployed positions of sensor nodes and communication signalattenuation, but also create the model of signal coverage to attain the optimal deployedlocations and the number of sensor nodes. These ameliorate communication coverage,improve the effective of the deployment and save the resources through probabilityoptimization. Through the simulations of different locating deployed strategies, relativedata proves the advantage and correctness of DDCP.(3) An effective algorithm based on signal coverage of effective communication isproposed on local energy-consumption saving strategy for the application in WSN.Hence, the multi-hop clustering routing algorithm of partition subspace is devised. Thisalgorithm fulfills local energy-balanced consumption ascribed to the topologicalacquisition of distributed sensor nodes locating deployment. However, the topology ofsensor nodes deployment is obtained by using gaussian distribution of the locating errormodel and the characteristics of markov chain, which improves the topologicalacquisition of the above algorithm. Through the simulation, the relative datademonstrate the advantages of our approaches on saving hardware resources and energyconsumption of networks.(4) In order to process the target tracking approximation with the unknown motionstate models and the measurement noise probability density function beforehand in atwo-dimensional field of binary proximity sensors, the algorithms based on costfunctions of particle filters and near-linear curve simple optimization are proposed.Through the moving target across the detecting intersecting fields of sensorssequentially, the algorithms make use of the geometrical features of sensing disks todetermine the limits of the boundaries of moving target and also the cost functions areintroduced to solve the target tracking approximation and controllable velocityestimation which are not similar to traditional particle filters that rely on probabilisticassumptions about the motion states. Then a near-linear curve geometric approach isused to simplify and easily describe the target trajectories that are below a certain errormeasure. The validity of our algorithms is demonstrated through simulation results. Atthe same time, the nodes’ failure detection mechanism is introduced to analyze the situations that some parts of sensor nodes are failure in practical applications. Theoriginal estimations confirmed by the failure sensor nodes are corrected depended onthe bayesian probability distribution which enhances the fault tolerance of the failuresensor nodes in event region, and also improve the precision of the target tracking.