期刊名称：Journal of Theoretical and Applied Information Technology
印刷版ISSN：1992-8645
电子版ISSN：1817-3195
出版年度：2013
卷号：49
期号：1
出版社：Journal of Theoretical and Applied
摘要：The recognition and localization of harmful acoustic signals in a certain scenario are of great importance. In this paper, we propose a scheme for harmful acoustic signals recognition and localization based on fish swarm algorithm for wireless sensor network. We firstly optimize the coverage where the harmful sound source can be detected efficiently. And then, we get the eigenvalue of the harmful acoustic signals by using the LPC (Linear Prediction Coefficients). After analyzing the similarity between the eigenvalue and the data pre-stored in the data base, we can find out the harmful sound. After detection, the harmful resource is located with the help of its closest nodes which are chosen based on time delay. Simulation results show that the proposed scheme can improve the efficiency of harmful source detection and accuracy of the localization.
关键词：Artificial fish swarm algorithm; Harmful acoustic signals; Linear Prediction Coefficients; Recognition; Localization
