摘要
由大量微小传感器节点组成的无线传感器网络主要用于从监测环境收集信息并做出相应的决策,研究与无线传感器网络相适应的战场目标识别算法具有重要的实际意义。针对无线传感器网络的自身特点和战场目标所辐射的声频率特性,提出了基于小波包和人工神经网络的无线传感器网络的声目标识别算法。利用小波包对声信号进行消噪处理并特征提取,然后利用神经网络分类器作最后的识别。最后实现了对该声目标识别算法的仿真。试验证明该声目标识别方法应用在无线传感器网络中是可行的。
The Wireless Sensor Network(WSN), composed by a great deal of micro-sensors, is mainly used to gather data information from the monitored environments and gives a decision-making. There is great practical significance in researching the arithmetic corresponding to battlefield target identification in WSN. In this paper, it comes up with a method for acoustic objectives identification, which is based on wavelet packet and artificial neural network (ANN). The arithmetic includes the wavelet packet to de-noise and extracts the feature of acoustic signals, then use artificial neural network to identify the objectives. We realize the WSN nodes' identification simulation for battlefield targets. The results indicate that this method is very feasible for the application in WSN.
出处
《电子测量技术》
2007年第12期63-65,69,共4页
Electronic Measurement Technology
关键词
无线传感器网络
小波包变换
特征提取
神经网络
wireless sensor network
wavelet packet transform
feature extraction
artificial neural network