Fast signal reconstruction and recognition algorithm based on cascading redundant dictionary and block sparsity for compressed sensing radar receiver

Abstract

The compressed sensing (CS) radar receiver based on the theory of CS has a great advantage in receiving wide-band signal and can receive high-frequency radar signal at a sampling rate far lower than the traditional receiver. However, in order to obtain the information of the target signal, the parameters of the reconstructed signal still need to be measured by conventional methods. In order to make full use of the compressed sampling and the characteristics of radar signals, this study proposed a cascading dictionary matching pursuit algorithm based on block-sparsity (BS-CDMP) to achieve fast identification and reconstruction of received signals. First, a signal-level redundant dictionary is constructed for quickly identifying the signal in which the authors are interested in. Then, they study the feasibility of cascading the signal-level redundant dictionary and the DFT basis as a whole sparsity dictionary. Finally, the block sparsity is introduced for the selection of different parts of the dictionary during greedy iteration. Experimental results show that their proposed algorithm has considerable and universal reconfiguration performance and can directly perform fast reconstruction and recognition of the target signal.

Keywords

• signal reconstruction
• radar receivers
• iterative methods
• compressed sensing
• radar signal processing
• signal sampling
• discrete Fourier transforms
• wide-band signal
• high-frequency radar signal
• traditional receiver
• target signal
• reconstructed signal
• compressed sampling
• cascading dictionary
• block-sparsity
• fast identification
• received signals
• signal-level redundant dictionary
• sparsity dictionary
• block sparsity
• fast reconstruction
• compressed sensing radar receiver
• DFT basis
• greedy iteration
• cascading dictionary matching pursuit algorithm