Abstract: In order to eliminate the influence of varying correlation of Wireless Sensor Network (WSN) data caused by transmission in the performance of the current Compressed Sensing (CS) reconstruction algorithms, a correlation adaptive reconstruction algorithm for network data was proposed. Firstly, the iterative algorithm was used to estimate the correlation of the date to be reconstructed, then two-step correlation of support set were used for coordinating the non-zero value in sparse coefficient vector, and eventually a more precise reconstruction of data was realized. The simulation result shows that this algorithm can effectively restrain the effect of noises in WSN data reconstruction and improve the accuracy of reconstruction under varying correlation.

Key words: Wireless Sensor Network (WSN), Compressed Sensing (CS), data reconstruction, correlation, adaptive

摘要： 为减小无线传感器(WSN)网络数据传输过程中相关性发生变化对压缩感知重构精度的影响，提出一种相关性自适应的网络数据重构方法。该方法首先通过迭代对待重构数据的相关性进行估计，进而采用支集元素的两步相关检验方法对网络数据稀疏系数向量中非零元素进行重构，最终得到更为精确的重构数据。仿真结果表明，该算法能有效抑制实际传输过程中各种干扰对网络数据重构的影响，提高网络数据相关性变化情况下的重构准确度。

关键词: 无线传感器网络, 压缩感知, 数据重构, 相关性, 自适应