基于CS和神经网络的传感器网络模式识别研究  

Study on sensor networks pattern recognition based on CS and neural network

作　　者：龚静[1] 陈向东[1] 时子青[1] 

机构地区：[1]西南交通大学信息科学与技术学院,四川成都610031

出　　处：《传感器与微系统》2011年第12期62-64,共3页Transducer and Microsystem Technologies

基　　金：国家自然科学基金资助项目(60871024);国家重点实验室开放课题项目(KFJJ200915)

摘　　要：当前,传感器网络快速地发展,其节点不断增多,需要传输和储存的数据量也就不断增大,同时现有传感器节点尺寸与复杂度限制了传感器网络的数据存储、数据的计算速度及频宽。针对此类问题,将压缩感知技术应用于传感器网络,对传感器网络中传输的数据进行压缩,降低传输数据量,然后在接收端重构数据,将该数据样本作为BP网络的输入进行识别。实验证明:该方法可以提高网络数据存储能力和带宽利用率。At present,with the rapid development of sensor networks and the increasing of nodes, the amount of data which are needed to transmit and store continue to increase. While the size and complexity of the existing sensor nodes limit the data storage, data calculation speed and bandwidth of sensor networks. Aiming at these problems, compressed sensing technology is applied to sensor networks, datas transmitted in sensor networks are compressed to reduce the amount of data transmission, and datas are reconstructed in receiving end, the data samples are recognized as the input of BP network. Experiment proves that this method can improve the capacity of data storage and the bandwidth utilization of network.

关 键 词：压缩感知 传感器网络 BP神经网络 

分 类 号：TP183[自动化与计算机技术—控制理论与控制工程]