基于无线传感器网络的声音目标定位算法

摘要

Maximum Likelihood Estimation is widely used in the multiple-source localization based on acoustic energy in wireless sensor networks; but it ignores the interaction effect of the targets' energy. This article used Independent Component Analysis to obtain the interaction effect of the targets' energy, established acoustic energy model on this basis, and at last estimated the targets' location using maximum likelihood method. The experiment result proves that this method delivers more accurate results and better anti-noise ability than the existing maximum likelihood multiple-source localization method.%无线传感器网络中基于声音能量的多目标定位常采用最大似然估计法,但该定位方法未考虑目标能量相互混合的影响.利用独立分量分析方法求得目标之间的能量相互混合的比例,在此基础上建立声音能量模型,并利用最大似然估计法对目标进行定位.实验结果表明,与已有的最大似然多目标定位方法相比,该方法具有更高的定位精度和抗噪性能.