传感通信网络过滤方法研究与仿真  

Research and Simulation of Filtering Methods in Sensor Communication Network

作　　者：纪松波[1] 白云[1] 

机构地区：[1]内蒙古工业大学信息学院,内蒙古呼和浩特010080

出　　处：《计算机仿真》2015年第2期328-331,共4页Computer Simulation

基　　金：内蒙古自然科学基金项目(2011MS0913)

摘　　要：传感通信网络中乘性噪声的过滤,可提高通信网络的性能。在传感通信网络乘性噪声过滤过程中,由于传感网络的信道大部分处于移动状态中,信道路线的变化会造成不同节点间的乘性噪声干扰,传统的传感网络方法过滤乘性噪声,信号与噪声之前缺少识别过程,统一转换成固定特征噪声加以删除,过滤效果较差。提出采用改进主分量分析算法的传感通信网络乘性噪声的过滤方法。对采集到的初始信号进行白化与降维处理,使初始有色信号转换成白色信号,且分量彼此间相互独立,减少处理过程的数据量,根据改进主分量分析算法构造信号数据矩阵,进行正交变换获取有效特征值并进行信号重构,实现对传感通信网络中乘性噪声的有效过滤。实验结果表明,利用改进算法进行传感通信网络乘性噪声的过滤,能够提高噪声过滤的精度,提高传感通信网络性能,具有极大的优越性。A multiplicative noise filtering method of sensor communication network is presented based on the im- proved principal component analysis algorithm. By bleaching and reducing dimension of collected initial signals, the initial colored signals are transformed into white, the component is independent each other, the amount of data pro- cessing is reduced, the signal data matrix is constructed based on improved principal component analysis algorithm, effective characteristic values are obtained by using orthogonal transform, the signals are reconstructed, and the mul- tiplicative noises in sensor networks are filtered effectively. The experimental results show that the improved algorithm can improve the accuracy of noise filtering and the performance of sensor network.

关 键 词：传感通信 乘性噪声过滤 主分量分析法 

分 类 号：TP391.9[自动化与计算机技术—计算机应用技术]