Target Source Localization Based On Wireless Sensor Network

Target location as an important application has a very important research significance in wireless sensor networks. Currently, There are three main methods of source localization: TDOA, DO A and received signal strength. The method of signal received strength based on need low energy consumption, simple location and less request on the synchronization of the network are widespread in the target monitoring environment. On the localization system, this thesis mainly studies from the following several aspects.Based on energy attenuation model, through the analysis on the comparing result of several energy intensity localization method, considering the complexity of calculation and size of energy consumption, a distributed localization scheme was proposed, and compared with the centralized algorithm. The results showed that relative to the distributed localization method, the centralized localization method had higher localization accuracy, but considering from the effect of energy saving, the distributed localization method had a better practicability.In the process of target localization the number of goal is not a single source in practice, this paper made simulation experiments through Gauss-Newton iterative algorithm in this case that there were two sound source targets in the sensing area. The simulation experiment results showed that the algorithm meets the need of accurate location for two source target. It is an important problem for the target located quickly and accurately under the condition that the sensor nodes are random distributed. The article converts acoustic target location problem into optimization of likelihood function based on RSSI of acoustic signal, then, it makes out the global optimum algorithm using DFP and BFGS method in optimization theory for target localization estimate. The simulation results showed that the method has good localization effect and application value.For the limited of resources energy and communication bandwidth in the wireless sensor network, It derived the likelihood function using energy attenuation model and gave the method of maximum likelihood based on acoustic energy.The simulation results show that the energy of network can be saved by reducing the amount of bits transmitted as long as location accuracy can be satisfied, And it also derived the Cramer-Rao lower bound of maximum likelihood estimator to evaluate the estimation result.