Multiple Gunshots ｓｏｕｒｃｅ Algorithm based on Integration of Cluster Analysis

Shu LI

Hunan City University, Yiyang, Hunan, 413000, China

Abstract: Gunshots positioning system can quickly determine the position of the gunshots ｓｏｕｒｃｅ. It will help to protect the safety of soldiers on the battlefield, assist police to crack down on violent fear of molecules. On how to improve the performance of multiple gunshots ｓｏｕｒｃｅ localization algorithm gunshots proposed multiple gunshots ｓｏｕｒｃｅ localization algorithm based on integration of sub frame estimation and cluster analysis. First, a signal is divided into 8 sub frames. Each sub frame signal using the phase variation is calculated in response to the power control weighting function, which searches the sound ｓｏｕｒｃｅ position to obtain the maximum value of the sub frame estimation. Because the signal in the time domain gunshot is sparse, these estimates correspond to a plurality of sound ｓｏｕｒｃｅ position. Then the estimated values of the sub frame is divided into several categories using the convergence of clustering algorithm. Finally controlled in response to the sub frame by the average estimated value of the power function was evaluated to give the final gunshot ｓｏｕｒｃｅ location estimation. Experimental results show that in the case of 2-3 gunshots, the positioning performance

of the algorithm is much higher than traditional algorithms.

Keywords: Gunshots Positioning System; Cluster Analysis; Sound ｓｏｕｒｃｅ Signal; Multiple Gunshots ｓｏｕｒｃｅ; Subframe; Anti-terror Attacks