基于压缩感知的无线传感网络数据压缩

无线传感器网络（Wireless Sensor Networks，WSN）负责感知、采集、处理和监控环境数据，但是容易受限于资源。压缩感知（Compressed Sensing，CS）理论表明，利用最优化理论，稀疏信号可以从少量的非自适应线性投影中高概率精确恢复。根据CS理论设计WSN的数据压缩方法只依赖于信号内在的结构和内容，而不是信号的带宽，弥补了WSN的不足；提出了基于稀疏随机投影的编码方法；仿真结果表明系统在满足误差要求条件下构造的数据包减少至结点数目的30%，提高了WSN通信效率，降低了系统能耗。

Data compression based on compressed sensing in Wireless Sensor Networks

Wireless Sensor Networks（WSN）are responsible for sensing, collecting, processing and monitoring environmental data, but resource is easily limited. The newly emerging Compressed Sensing（CS）theory holds that sparse signals can be exactly reconstructed with high probability from a small amount of non-adaptive linear measurement through optimization. A data compression method which is dependent only on the structure and content of the signal, rather than the bandwidth of the signal is designed by sparse random projections through network coding. The results of simulation show that this system not only improves the efficiency of WSN communication by reducing packets to 30% number of nodes, but also reduces the system energy consumption under the error requirement.