迭代维纳滤波在水声信号分离中的应用

Application of Iterative Wiener Filtering for Sonar Signal Separation

论文对迭代维纳滤波在水声信号分离中的应用进行了研究,并根据特定目标声源信号的听觉特征先验知识,采用水声信号的听觉特征参数匹配度作为迭代收敛判据。迭代滤波能够逐渐消除干扰声源信号,模拟人类听觉系统的注意力机制,使滤波结果的听觉特征最大程度地与目标声源匹配。仿真试验结果表明,这种迭代维纳滤波方法在有干扰声源存在的情况下可以有效实现水声信号分离。

Iterat ive wiener filtering for sonar signal separation is researched in this paper. Based on the prior knowledge of the target acoustic signal’s auditory feature, the iterate is controlled by the matching degree of sonar signal auditory feature parameters. Iterative filtering attenuates interference signal continually to simulate the attention of human auditory system, and makes the matching degree between filtering output and target source signal maxim. Simulation result shows that the iterative filtering can separate sonar signal efficiently in the interference source signal environment.