摘要
在无线传感器网络(WSN)中,以往都是采用奈奎斯特技术对信号进行采样并重构,而随着信号频率的增加,应用奈奎斯特技术会使成本大幅度的增加,这是人们所不乐见的。针对这一问题,近年来出现一种新的技术即压缩感知技术,它能利用更少的数据和合适的重构方法得到更精确的原始信号。将稀疏贝叶斯学习(SBL)和压缩感知联合起来,形成了一种在有噪声的情况下更好重建可压缩信号的方法,并进一步将这种方法应用在WSN中,可以在误差允许的范围内有效控制测量数据的维数,在保证一定误差的同时还减少了成本,提高了算法的效率。
In wireless sensor networks,signal is sampled and reconstructed using the technology of Nyquist in the past. But it requires a substantial increase in the cost with the growth of the signal frequency,which is that people do not like to see. Recently a new technology is emerged,which is called compressive sensing technology. Compressive sensing can use less data and appropriate reconstruction method to get a more accurate original signal. Put Sparse Bayesian Learning （ SBL） and compressive sensing together to form a better way of re-constructing compressible signal under the noise. This method can effectively control the dimension of measurement data within the range of allowed error in WSN,so you can ensure a certain degree of error while reducing the cost,improving the efficiency of the algorithm.
出处
《计算机技术与发展》
2014年第3期198-201,共4页
Computer Technology and Development
基金
国家自然科学基金资助项目(60972041
60972045)
关键词
无线传感网络
压缩感知
贝叶斯模型
信号重构
wireless sensor networks
compressive sensing
Bayesian model
signal reconstruction
