The Research Of Underwater Sensor Networks Deployment Based On Data Fusion

Underwater sensor networks, as an important branch of sensor network research, has a broad application prospect in many fields, such as, in marine data monitoring, ocean flow environmental monitoring, ocean disaster prevention, and underwater ships navigation, survey of Marine resources, distributed monitoring of the tactics and underwater target detection, tracking and positioning has extensive application prospect. This subject in view of node deployment issues of the underwater sensor network. The problems of underwater sensor network node deployment is studied on the basis of the existing evaluation indexes, perception model, data fusion, deployment algorithm, etc.For 3D nodes-deployment of the underwater sensor networks, we proposed definitions of detection reliability and reliability coverage. By the decision criterion of advanced D-S theory of evidence, we analyzed the changes of reliability coverage in underwater sensor networks’ detection area. According to the results of analysis and combined with the prior probability of target nodes distribution, we proposed a Nodesdeployment Algorithm Based on Advanced D-S Theory of Evidence and Prior Probability(NAAEP). Simulation results demonstrated that the algorithm can effectively reduce the number and energy cost of sensor nodes, expand detection area, improve the detection performance of sensor networks.Based on the probability sensing model of passive sonar, we established fuzzy sensing model and fuzzy data fusion model. We also studied the importance of fuzzy data fusion for the nodes-deployment of sensor networks, proposed nodes-deployment strategy based on fuzzy data fusion for underwater sensor networks(NAFC). The effectiveness of NAFC is proved using MATLAB, energy consumption and the number of deployment nodes is reduced, the lifetime of network is prolonged, and the scope of perceiving is expanded. As a result, NAFC can improve the detection performance of the sensor network.For node-deployment of heterogeneous sensor networks, based on different probability sensing models of heterogeneous nodes, refer to organic small molecule model, we proposed ‘class-molecule’ sensing model of heterogeneous nodes. Combined with DSm T data fusion model, we studied the changes of network coverage after using the new sensing model and data fusion. According to the research results, the node-deployment strategy of heterogeneous sensor networks based on organic small molecule model(NHOS) is proposed. Sensor network simulation model is established using MATLAB, the simulation results show the effectiveness of the algorithm, the coverage of network and detection efficiency of nodes are improved, the lifetime of network is prolonged, energy consumption and the number of deployment nodes is reduced, and the scope of perceiving is expanded. As a result, NHOS can improve the detection performance of the network.