The Design Of Low-complexity Iterative Receiver For Polar Coded Sparse Code Multiple Access System

The requirements of the fifth generation(5G)mobile communication system motivate the novel wireless communication techniques to be more accurate,more efficient and more energysaving.Compared with the forth generation(4G)mobile communication system,the 5G system has significant improvements on transmission rate,spectral efficiency,energy efficiency and mobility.In the 5G protocol framework,three typical usage scenarios are considered in order to meet special requirements for different situations,they are: enhanced mobile broadband(eMBB),massive Machine Type Communications(mMTC)and Ultra-Reliable and Low Latency Communications(URLLC).In the point of obtaining the high-quality communications,non-orthogonal multiple access(NOMA)becomes a potential multiplexing technique for mMTC and URLLC due to its flexible carrier configurations,massive connectivity and high spectral efficiency.Meanwhile,sparse code multiple access(SCMA)is a competitive candidate among several NOMA techniques.In order to support control sequence transmission and grantless transmission in 5G,polar code is voted as the channel coding scheme for eMBB control channel and several broadcast channels with short-length packet,which are resulted from its outstanding performance of short-length codes and low-complexity decoding algorithm.This thesis studies the iterative receiver,in which polar codes and SCMA are employed as channel coding and multiple accessing schemes respectively.The first chapter of the thesis concerned the backgrounds and motivations of the designed iterative receiver of the polar coded SCMA system,which includes the importance of key capabilities of the 5G networks,the 5G protocol framework of physical layer and the current status of NOMA and polar codes according to 5G.Based on this,we further extract the main problem that should be solved in this thesis,that is the effective design and analysis of an iterative polar-SCMA receiver in the case of being short of high-powered receiving strategies that under the specifications of the 5G physical layer.The second chapter proposed a low-complexity detecting scheme based on the partial marginalization(PM).Such detecting scheme is mainly applied to uplink SCMA system for multi-dimensional detection.We combined the message-passing algorithm(MPA)of SCMA with the idea of PM,which is further enforced to decrease the computational complexity(CC)of MPA.The proposed PM-MPA detector categorizes the received signals and disposes them block-wisely during the updating processes between source nodes and variable nodes.The outputs of the variable nodes of PM-MPA are transformed from the product of entire iterative information(EII)of MPA into the product of EII and partial iterative information(PII)by userdefined parameters.In which,the CC of PII is much less than that of EII,thus,the whole CC of PM-MPA decreases from MPA significantly.Finally,simulation results proved the effectiveness of the proposed PM-MPA by investigating the CC,bit error rate(BER)and convergencebehavior.The third chapter proposed an iterative detection-decoding scheme for short-length polar coded SCMA system,where successive cancellation(SC)decoder is employed.Because of the SC decoder is unable of generating the entire soft information of polar codewords,the core of the proposed scheme is to complete the feedback procedure by designing a re-encoder that is able to compile the entire soft information of polar codewords.The re-encoder is equipped after SC decoder,which produces the soft information by the use of decoded information.Additionally,the re-encoding complexity is decreased by summarizing the essential elements of polar encoding and taking full advantages of the characters of the generating matrix and frozen bits.Then,an iterative SC polar decoder is provided by constituting the SC decoder,the re-encoder and the hard-decision unit altogether.Such iterative decoder is able to generate soft information of all polar codewords and send them back to the detector as priori information.Finally,the whole iterative receiver is realized by the combination of the iterative decoder and the detector,without clipping,amplification and hard-decision during the iterations,coding gain and iterative gain are obtained.Simulation and numerical results proved the feasibility and availability.The forth chapter proposed an iterative detection-decoding scheme for long-length polar coded single-user system.In this chapter,we replace the SC decoder by the successive cancellation list(SCL)decoder to further improve error correcting capability of the system,where SCL decoding processes are elaborated in the first section.To solve the problem of high computational complexity of the soft information that resulted from the re-encoding of long-length polar codewords,we provide two modified re-encoders on the basis of the original re-encoder that has been described in last chapter,named Max-Log-based re-encoder(ML-R)and Min-Sumbased re-encoder(MS-R).Both ML-R and MS-R are able to be employed in different scenarios for special requirements,the trade-off between re-encoding complexity and re-encoding accuracy can also be made by configuring user parameters.Simulation results show that the BER performance of MS-R is slightly better than ML-R in the priority-free scenario,and under such circumstance,these two re-encoders are suitable for the iterative polar coded receiving strategy where SCL decoder is employed.In last chapter,based on the construction of long-length polar coded multiuser SCMA system,the effects of iterative receiving on mutual information are studied,and the layering phenomenon of mutual information of the polarized sub-channels are also investigated.For extensive researching,the CC of the proposed iterative receiving strategy of polar codes is compared with that of the existed Turbo and LDPC codes.Besides,in this chapter,we systematically derived the equivalent relationship between B-DMC and BSC of the mutual information under the condition of binary input for polar codes.Then,from the point of signal detection,the mutual information layering phenomenon of sub-channels,which caused by the proposed iterative receiving,is explicitly analyzed,where transmitted bits are mapped into multidimensionalSCMA symbols.Finally,the polar coded iterative receiving strategy shows the advantages in the aspects of receiving complexity and BER/BLER performance compared with other codes,especially in line with the configurations of the 5G physical layer.After all,the proposed iterative receiving strategy of polar codes is proven to be closer to the Shannon limit according to average mutual information(AMI).