Design Of Spatially Coupled LDPC Codes For NOMA System

Low Density Partiy Check(LDPC)codes attract much attention in the field of communication research due to their excellent performance.Spatially coupled LDPC(SC-LDPC)codes,a particularly important class of LDPC codes,combine the best features of optimized irregular and regular codes in one design: capacity approaching iterative belief propagation(BP)decoding thresholds and linear growth of minimum distance with block length.Therefore,SC-LDPC codes have been recognized as a highly promising channel coding scheme.Since SC-LDPC codes have a convolutional structure,they are expected to have a good performance over block-fading channels.On the other hand,non-orthogonal multiple access(NOMA)has been recognized as a promising multiple access technique for the wireless networks due to its high spectral efficiency,low latency,massive connectivity,and high user fairness.Under the assumption of perfect successive interference cancellation(SIC),NOMA with capacity achieving channel coding has been shown to approach the capacity bound of multi-user systems.However,practical wireless systems always use finite-length channel codes,causing decoding failure.Besides,the imperfect SIC introduces severe error propagation into NOMA systems and accordingly detriments the system performance.Therefore,the channel code design under NOMA system needs to be further studied.In this thesis,the serial connection multi-SC-LDPC(SC-MSC-LDPC)codes for blockfading channels and NOMA systems are studied.The specific contents are given as follows.First,we studied the construction and structural characteristics of SC-LDPC codes from the perspective of the protograph and the base matrix.In order to deepen the understanding of block-fading channels,we also review the characteristics of block fading channel model,the definition of the outage probability and the Protograph-based EXIT(PEXIT)algorithm.Secondly,the construction of SC-MSC-LDPC codes is studied for the block-fading channel,and the coding scheme is extended for different types of SC-LDPC codes.Multiple sub-chains are coupled by exchanging the edge connections of variable nodes at the same position of each sub-chain to strengthen the correlation between different fading blocks.There is no need to add additional edges or nodes,and the SC-MSC-LDPC codes would inherit the degree distribution of the sub-chains.Based on the PEXIT algorithm,it is proved that the infinite-length performance of the SC-MSC-LDPC codes can approach the outage limit.And the performance of the SC-MSC-LDPC codes are analyzed by the belief propagation(BP)decoding algorithm.Simulation results show that the SC-MSC-LDPC codes can improve the performance of the single-chain SC-LDPC codes and achieve full diversity.In order to improve the error performance,a particular class of(3,6,L)SC-MFSC-LDPC codes constructed by the folded SC-LDPC codes is proposed.Simulation results show that the optimized(3,6,L)SC-MFSC-LDPC codes can acquire a better coding gain.Finally,the performance of SC-MSC-LDPC coded downlink NOMA system is studied,with the emphasis on the SIC detection performance of the strong user.The performance of the strong user under different power distribution coefficients is analyzed and the simulation result shows that the decoding performance of the strong user will gradually converge when the power distribution coefficient is less than 0.25.And the comparison of the actual SIC detection with the ideal SIC detection demonstrates that the proposed codes can obtain excellent performance in the downlink NOMA system.