The Research And Modeling Of Source Localization Based On Improved Gauss Newton Iterative Method

Wireless sound sensor has been widely used in noise suppression, speech recognition, tracking target sound source and so on. In the daily life, the wireless sound sensor network has been widely used in the sound source localization, for example, gunfire positioning, video conferencing and so on. In order to accurately locate the sound source and study the error of the results, in this paper we adopt the sound source localization method based on wireless sensor network(WSN), using time difference of arrival method to locate the sound source in or around the triangular array. And we model and simulate the sound source location system that contains different wireless sensors based on the proposed method and then we analysis the effect of node location error and sound arrival time error.When we apply the Gauss Newton iterative method to the sound source localization, the initial value of the first iteration has a great influence for the result of the algorithm. the inappropriate initial value will lead to the wrong result which may cause inaccurate results. Hence, this paper proposes an improved Gauss Newton iteration method aiming to select the good initial value. Dividing the area of the wireless sensors, we choose the appropriate initial value by the improved Gauss Newton iteration method. The method can improve the accuracy of the calculation results, and reduce iteration times and error during the calculation process.The simulation results show that the effective location range of the sound source localization system based on the wireless sensor network is mainly located in the triangle area, and has certain extension to the outside of the triangle area. With the expansion of the area of the sound sensor array, the effective locating range is also enlarging. The sound source location system has a larger effective locating range when the sensor array is aligned into a positive triangle with the similar array size and the other error condition. The simulation results also show that in the case of the error of the system, when the error smaller, the positioning effect is more ideal.And the influence of the position error of the sound sensor node on the locating results is greater than the impact of the sound arrival time error.