Research On Target Tracking Algorithm Of Underwater Wireless Sensor Network

Underwater wireless sensor network is composed of a large number of sensor nodes with the limited capacity of detection,sensing,communication and computation.With the exhaustion of land resources,human has started shifting their focus to the ocean.As a significant component of ocean exploration and maritime security,underwater target tracking has attracted extensive attention.Underwater wireless sensor network mainly adopts acoustic communication,which makes the communication bandwidth much more limited than that of terrestrial wireless sensor network.In addition,the energy stored by each sensor node is limited and the underwater wireless sensor network is usually deployed in a wide range of three-dimensional space,which makes it difficult to charge or replace the battery once the node runs out of energy.Therefore,only part of sensors can participate in tracking the target at each time step,so as to save energy.The application of the target tracking algorithm must consider the characteristics of actual application scenarios,such as measurement environment and resource constraints.From the perspective of the limited energy and limited communication bandwidth of the underwater wireless sensor network,this thesis investigates tracking algorithms to maximize the information benefit under the condition of using the same resources.In the thesis,the network structure of underwater wireless sensor network,the motion model of target tracking and common filtering algorithm are introduced.Then the usability of different filtering algorithms in different scenarios are analyzed.Finally,the difficulties and challenges of underwater wireless sensor network target tracking are clarified.The two algorithms proposed in this thesis solve the problem of limited energy and communication respectively.In addition,in view of the non-linear propagation characteristics of underwater acoustic,the measurement model with multiplicative noise is established to reduce the measurement error caused by it.From the perspective of limited energy,only part of the sensors can participate in the tracking at each time step.The topology of the sensors relative to the target has a significant impact on the estimation results.Based on the fact that each sensor node can easily adjust its own depth with the sound pressure sensor,an optimal sensor depth adjustment algorithm is proposed,which can improve the topology structure between the node and the target through depth adjustment.The tracking performance can beimproved under the condition of selecting the same number of sensors to participate in the tracking.However,this algorithm is based on the original measurement.Considering the limited bandwidth of underwater communication,each sensor needs to quantify the measurements detected by itself before sending them to the fusion center.Therefore,an optimal quantization bit allocation algorithm is proposed to maximize the contribution of the data under transmission bits constraint.At the same time,a genetic algorithm for bit allocation is proposed to improve the real-time performance of the algorithm.