A fast algorithm of compressed sensing

Author

Hao Liu ; Yuehao Yan ; Junhai Wang

Author_Institution

Dept. of Comput. Sci., Sichuan TOP Vocational Inst. of Inf. Technol., Chengdu, China

fYear

2014

Firstpage

463

Lastpage

467

Abstract

After studying the compressed sensing theory and its main reconstruction algorithm-Matching Pursuit (MP) algorithm, this paper proposes a new approach to improve the speed of MP algorithm, and it describes how to build a Beowulf parallel computing system with 8 PCs. Its parallel computations is implemented by Message-Passing-Interface(MPI), and a 100Mb/s high speed Ethernet network interconnects all PCs. Test is made using parallel computing program to measure the parallel efficiency of the system, results show that this approach can reduce the MP algorithm computing time-cost form 78 minutes with a PC to 11 minutes with 8 PCs.

Keywords

application program interfaces; compressed sensing; iterative methods; message passing; parallel algorithms; parallel programming; time-frequency analysis; workstation clusters; Beowulf parallel computing system; Ethernet network; MP algorithm; MPI; bit rate 100 Mbit/s; compressed sensing; fast algorithm; matching pursuit algorithm; message-passing-interface; parallel computing program; Compressed sensing; Computational modeling; Computers; Dictionaries; Matching pursuit algorithms; Parallel processing; Program processors; Compressive sensing; MPI message passing; Matching Pursuit (MP); parallel computing;

fLanguage

English

Publisher

ieee

Conference_Titel

Wavelet Active Media Technology and Information Processing (ICCWAMTIP), 2014 11th International Computer Conference on

Print_ISBN

978-1-4799-7207-4

Type

conf

DOI

10.1109/ICCWAMTIP.2014.7073450

Filename

7073450