Node Self-localization And Passive Acoustic Source Target Positioning Based On Micphones Of Smartphones

In the field of wireless sensor networks, node self-localization and target localiza-tion have become a research hotspot. Based on the non-cooperative target characteris-tics, passive acoustic source targeting technology can achieve high precise positioning, which has been applied widely in the military and civilian fields. The current main-stream technology has disadvantages such as high cost of the customized hardware module, limited endurance capacity, uncontrollability of synchronization accuracy, technologies supported by generic smart phone could only localize the acoustic source under friendly and cooperative circumstance, technologies supported by the custom-ized hardware module can’t avoid the time delay uncertainty between the wireless synchronization of nodes and the acquisition of acoustic signals.In view of the above problems, our method for acoustic source location based on TDOA is supported by universal smart phone hardware, which is suitable for not only the active target acoustic source, but also the passive acoustic target source, and which uses only acoustic signal which make the implement simplify. The audible signals shorten the need of customized ultrasonic excitation circuit module or other hardware platforms, which are convenient in use and cost-effective. The main work is as fol-lows:Firstly, choosing the suppress of the multipath effect as the breakthrough point, the acoustic signal delay estimation method based on mutual sparse is proposed in that the threshold parameter does not have the adaptability during suppressing the multipath effect when the delay estimation method based on generalized cross corre-lation is used, which can effectively suppress the negative effects of multipath effects by sparsely reconstructing cross-correlation vector. This method mainly uses to obtain TDOA time delay estimation for passive acoustic signal and TOA time delay estima-tion for chirp sound signal. The simulation results show that the proposed method is of high accuracy and wide application range.Secondly, in view of classic Multidimensional Scaling Location Algorithm on practicality and reliability is insufficient, a weighted multidimensional scaling self- localization algorithm based on TPSN model (T-WMDS) for sensor network node is proposed. This algorithm is to serve in the smart phone platform to achieve the acous-tic source passive localization. In the simulation, we evaluate the performance of the new algorithm under the change of the number of beacon and the background noise. The result shows that the localization accuracy of the proposed algorithm is close to the CRLB when the noise variance is enough small.Finally, we introduce the design of passive acoustic target localization system and the implement on the smart phone platform. For the stationary acoustic source target, the system introduced the combination of statistical decision and Voice Activity De-tection (VAD) method to estimate time delay. For the slowly moving low altitude acoustic source target, the system use sparse cross-correlation method to estimate time delay. Aiming at two different types of passive acoustic source target (peccancy whis-tling car and slowly moving low altitude four rotor unmanned aerial vehicle), the ex-perimental scenarios and solutions are designed and the results are analyzed. In addi-tion, the self-localization algorithm we proposed is also verified from the perspective of experiment.