文献类型：会议论文

英文题名：Fast sparse compressive sensing based on the wavelet package hierarchical trees

作者：Chen, J.X.[1]; Wang, T.M.[1]; Niu, W.L.[1]; Du, L.J.[1]

第一作者：陈景霞

机构：[1] College of Applied Science and Technology, Beijing Union University, Beijing, China

第一机构：北京联合大学应用科技学院

会议论文集：Network Security and Communication Engineering - Proceedings of the 2014 International Conference on Network Security and Communication Engineering, NSCE 2014

会议日期：December 25, 2014 - December 26, 2014

会议地点：Hong Kong, China

语种：英文

外文关键词：Bandwidth compression - Compressed sensing - Data compression - Forestry - Pipelines - Signal reconstruction - Signal to noise ratio - Wavelet analysis

摘要：For efficient and consistent signal compressive sampling and reconstruction, with the analysis of the sparsity of mutation smooth signal and its structural in wavelet space, a new Fast Sparse Compressive Sensing (FSCS) method, which decomposes signal with the rule of fast sparse based on its structural characteristics, was developed based on the multi-level tree model of wavelet packet. Combined with the established reconstruction algorithm, the effectiveness of the proposed Compressive Sensing method was evaluated with typical industrial mutation smooth signals, e.g., the pipeline leakage signal. With the acceptable root-meansquare error and signal to noise ratio of reconstructed signal, the compression ratio increased significantly, which was helpful to reduce the burden of communication bandwidth of industrial monitoring networks. With the compression ratio of 30:1, the parameter of ERP of the developed method is less than 1×10–3, which ensures the accuracy of reconstructed signal for pipeline leakage detection and positioning. ? 2015 Taylor & Francis Group, London.