Distributed Compressed Sensing Based on Local Transformer Network
Yu Zhou,Weixin Xie,Yong Yang,Shilong Sun,Xiao Zhang,Jun Qin
DOI: https://doi.org/10.2139/ssrn.4630798
2023-01-01
Abstract:Among compressed sensing algorithms, Distributed Compressed Sensing (DCS) stands out for its ability to jointly reconstruct multiple signals, making it a focal point of interest for researchers and engineers , thanks to its outstanding efficiency in signal reconstruction. In contrast to conventional approaches that depend on a priori estimates, like joint sparsity of signals, deep neural networks leverage their robust feature extraction and sequence modeling abilities to automatically capture inter-signal dependencies. This confers them with the benefits of faster solving speed and superior reconstruction accuracy. Most of the current methods, however, do not fully utilize the global dependency information between signals, while also ignoring the detailed features within the signals. Therefore, this paper proposes a Local Transformer Network for DCS (LTN-DCS). It utilizes the self-attention mechanism to extract global contextual information between the reconstructed signal sequences, while also addressing local redundancy partitioning within the signal sequences for finer-grained information capture. This approach generates probabilistic estimates of the non-zero element positions in each signal, forming the foundation for selecting basis atoms from the sampling matrix. Finally, the least-squares method is used to achieve the signal reconstruction. Through extensive experiments, the results show that our proposed method outperforms state-ofthe art DCS methods.