Compressed sensing‐based unequal error protection by linear codes. Issue 7 (1st September 2014)

• Record Type: Journal Article
• Authors: Li, Weiwei; Jiang, Ting; Wang, Ning
• Is Part Of: IET signal processing. Volume 8:Issue 7(2014)
• Page Start: 800
• Page End: 808
• Publication Date: 2014-09-01
• Subjects: compressed sensing -- linear codes -- channel coding -- linear programming -- minimisation -- signal reconstruction -- wireless channels -- discrete wavelet transforms; compressed sensing‐based unequal error protection -- linear codes -- wireless communication systems -- UEP techniques -- linear programming -- discrete wavelet transform -- DWT‐coded information -- time‐varying characteristic -- corrupted channel -- transmission scheme -- error‐free transmission -- l1 ‐minimisation algorithm -- signal reconstruction -- peak signal‐to‐noise ratio -- error resilience -- equal error protection scheme -- channel corrupted ratio -- PSNR drop rate; Signal processing -- Periodicals; 621.3822
• DOI: 10.1049/iet-spr.2013.0366
• Languages: English
• ISSNs: 1751-9675

Abstract

Abstract : In many wireless communication systems, data can be divided into different importance levels. For these systems, unequal error protection (UEP) techniques are used to ensure lower bit error rate for the more important classes. Moreover, if the precise characteristics of the channel are known, UEP can be used to correctly recover the more important classes even under severe receiving conditions. In this study, a UEP scheme based on compressed sensing via a linear program is proposed. Discrete wavelet transform (DWT) is chosen as the sparsifying basis, and then DWT‐coded information is divided into two‐layered coded streams, each of which is transmitted differentially by applying an unequal number of information bits in linear codes according to the time‐varying characteristic of the corrupted channel. In this proposed transmission scheme, the more important information is to guarantee error‐free transmission. At the decoder, one can simply reconstruct the signal via the l 1 ‐minimisation algorithm. Simulation results show that the proposed scheme can achieve a higher peak signal‐to‐noise ratio (PSNR) and obviously improve the error resilience compared to the equal error protection scheme and other UEP methods. More importantly, with the increase of channel corrupted ratio, the drop rate of PSNR is much slower than other solutions. It indicates that the proposed method has better robustness for severe channel conditions.