Research On Acoustic Source Localization Using Dual Sensing Nodes In Wireless Sensor Networks

The rapid development of computer technology has driven the wide development space and utilization of acoustic source localization techniques,acoustic source localization can be used in many occasions including locating the speaker in the video conference,locating the explosion source,and shooter localization.However,most existing localization technologies have problems of low computing capacity,high system costs,and low positioning accuracy in the presence of reverberation environment.To reduce the influence of all above disadvantages,this thesis proposes four novel acoustic source localization schemes based on dual sensing nodes in wireless sensor networks.This thesis proposes two acoustic source localization methods based on dual-microphone nodes in wireless sensor networks,and explores the underwater acoustic sensor networks and proposes two error tolerant dual-hydrophone source localization methods.Above all proposed methods utilize sensor nodes equipped with pairwise microphone or dual-hydrophone,which can remove the time-synchronized requirements.They use the sign of TDOA to convert the localization problem into plane cutting issues and half-plane intersection problem,which can reduce the demand of GCC algorithm,and reduce the hardware requirements of the system.In addition,utilizing probability to determine the binary information can reduce the positioning error caused by uncertain reasons or node measurement error.By narrowing the feasible region of the source,two methods finally use the peak of weighted probabilities or Monte Carlo algorithm to estimate the source's location.The thesis conducts simulation in different aspects for all proposed methods to evaluate their localization performance,and compares them with SO-TDOA based localization method and traditional hyperbolic intersection based TDOA localization method.The results indicate that these methods can locate the source,and achieve highly robustness when angle,measurement and location error exist,and when there exist error,their localization performance is better than that of TDOA-based method and SO-TDOA based method.