Distributed Compressive Sensing for Wireless Sensor Networks
In industrial applications of wireless sensor networks (WSNs), due to the constraints of limited bandwidth and power, it is not easy to transmit vibration signals in a real-time manner. To address this problem, compressive sensing is exploited to perform vibration signal compression with low computational complexity, which reduces volume of transmitted data and increase transmission efficiency and real-time performance. The fundamentals of compressive sensing enabled signal communication include exploring the inherent sparsity of vibration signals in the frequency domain, compressing the signal with a random Gaussian sampling matrix, and reconstructing the original signal by L_1 norm optimization. A distributed computing paradigm is also proposed to improve the performance of reconstruction by fusing the information of multiple sensor nodes. Experiments show that the proposed distributed compressive sensing enabled signal communication can decrease computational load, improve real-time performance and satisfy the constraints, making it a promising technology for communication in WSNs.
Wireless sensor networks compressive sensing sparsity Gaussian random matrix
Sun Xinyao Wang Xue Wang Sheng Bi Daowei
State Key Laboratory of Precision Measurement Technology,Tsinghua University,Beijing 100084 China
国际会议
厦门
英文
513-519
2010-05-22（万方平台首次上网日期，不代表论文的发表时间）