压缩感知在无线传感器网络数据采集中的应用

提出了一种无线传感器网络中基于压缩感知的数据采集方法。通过分析信号压缩观测过程,提出了适合在硬件资源有限的传感器节点中实现的循环稀疏伯努利观测矩阵CSBM(Cyclic-Sparse-Bernoulli Measurement),该矩阵使用循环稀疏矩阵与伪随机伯努利序列,采用结构化的方法构造,具有非零元素少、良好的伪随机性、硬件易于实现等优点。仿真实验表明,与其他类型的观测矩阵相比,CSBM矩阵在一定信号重构精度前提下具有更低的压缩采样比CSR(Compress Sampling Rate)。在无线传感器网络数据采集应用中,感知节点可以通过压缩观测得到更少的观测数据,能够大大减少网络通信数据量。

Compressed Sensing for Data Collection in Wireless Sensor Network

An efficient data collection method in wireless sensor network(WSN) was proposed based on Compressed Sensing(CS). Firstly, the Cyclic-Sparse-Bernoulli Measurement(CSBM) matrix was presented which is suitable for application in resource-constrained sensor node. The CSBM matrix is constructed using the structured approach with cyclic-sparse matrix and Bernoulli pseudo-randomness sequence, which have a series advantages, such as less nonzero elements, good property of pseudo-randomness and easy implementation in hardware. The Simulation and experiment show that considering the precision of signal reconstruction, the CSBM matrix can reach the less compress sampling rate(CSR) compared to other types of measurement matrix. In the application of data collection in WSN, the sensor node can acquire less data through CS measurement, which reduce data traffic in WSN.