A Sparse Binary Coding Circuit for Compressed Sensing

Abstract

The invention relates to a sparse binary encoding circuit for compressed sensing. A normal phase input end of a numeric comparator is connected with an output end of a linear feedback shifting register; a normal phase input end of a subtracter is connected with the output end of the linear feedback shifting register; an input end of an either-or data selector is respectively connected with the output ends of the linear feedback shifting register, the numeric comparator and the subtracter; an output end of the either-or data selector is connected with an input end of a random access memory; anoutput end of the random access memory is connected with an input end of an adder; and an output end of the adder is connected with the input end of an access storage. A measurement matrix corresponding to the sparse binary encoding circuit disclosed by the invention has high sparsity; each column only has one value of 1; an accumulating operation is only carried out for once in each clock period;calculation complexity is low; and the sparse binary encoding circuit is convenient for hardware implementation. According to the invention, both a circuit area and power consumption of the sparse binary encoding circuit are reduced to a great degree relative to a conventional binary encoding circuit, so that system power consumption is reduced, and system continuation is promoted.

Description

technical field [0001] The invention relates to an encoding circuit. In particular, it relates to a sparse binary coding circuit for compressed sensing. Background technique [0002] Natural information usually contains a lot of redundancy and has high compressibility. In the traditional processing method, the information is completely collected according to the Nyquist sampling theorem in the analog-to-digital conversion stage, and then input to the digital system for compression, and then the compressed result is for transmission or storage. However, this method has certain defects, that is, the intermediate link contains a large amount of redundant information processing, and an additional digital system needs to be introduced for compression coding, which increases the cost of hardware implementation. [0003] Candes and Donoho et al. proposed the theory of compressed sensing. Using the sparsity of the signal, the signal can be sampled far below the Nyquist frequency a...

AI Technical Summary

Problems solved by technology

The observation matrices corresponding to the above encoding circuits are all dense matrices, with high computational complexity, inconvenient storage and transmission, complex circuit structure, large area and power consumption, which are not conducive to integration

Images