Authors: Jingyi Wang; Shuling Che; Ying Li; Junlong Wang
Addresses: State Key Laboratory of Integrated Service Networks (ISN), Xidian University, Xi'an 710071, China ' State Key Laboratory of Integrated Service Networks (ISN), Xidian University, Xi'an 710071, China ' State Key Laboratory of Integrated Service Networks (ISN), Xidian University, Xi'an 710071, China ' Huawei Technologies Co. Ltd., Xi'an 710071, China
Abstract: Optimisation of degree distribution for joint network LDPC code over orthogonal Multiple-Access Relay Channels (MARC) is investigated in this paper. We first use a differential evolution algorithm to optimise the degree distribution of the joint network LDPC code over orthogonal MARC based on Gaussian approximation. Then, using a modified differential evolution algorithm, we design a rate-compatible network LDPC code, which is equivalent to designing a single code with different Signal-to-Noise Ratio (SNR). Simulation results show that the gap between Bit Error Rate (BER) performance of the former algorithm and the decoding threshold is about 0.4 dB and that between the capacity limit and decoding threshold is less than 0.7 dB at BER = 10−6. Simultaneously, the gap between BER performance of the latter algorithm and the decoding threshold is about 0.52 dB and that between the capacity limit and the decoding threshold is less than 0.46 dB at BER = 10−7.
Keywords: network LDPC codes; low-density parity-check codes; multiple-access relay networks; Gaussian approximation; differential evolution; capacity limit; optimal design; degree distribution; orthogonal MARC; multiple-access relay channels; signal-to-noise ratio; SNR; simulation; bit error rate; BER performance; decoding threshold.
International Journal of Grid and Utility Computing, 2016 Vol.7 No.1, pp.68 - 74
Available online: 18 Dec 2015 *Full-text access for editors Access for subscribers Purchase this article Comment on this article