Abstract:

Aiming at the change of sparse structure introduced by mobility of the wireless sensor network（WSN） nodes and noise in data transmission, an adaptive sparse transform method based on dictionary learning （DL）for WSN data was proposed. The optimum sparse basis can be adaptively constructed according to the change of sparse structure, and the compressibility of WSN data basis was introduced to DL to satisfy the real time requirement for large-scale data processing. Analysis and experimental results demonstrate that the proposed algorithm can significantly improve the robustness and the real time performance of WSN data sparse transform.

Key words: compressed sensing, dictionary learning, wireless sensor network, sparse transform, compressibility