Research On The Design Principles Of Joint Coding And Lossy Source Coding System Based On P-LDPC Codes

In the 5th generation mobile networks(5G)era,some applications will become popular,such as driverless cars,drones,and smart cities.In the 6th generation mobile networks(6G)era,these applications will become more powerful.This puts forward higher requirements for communication technology,and communication system design faces more serious challenges:low energy consumption will become more important.Therefore,the development of 6G requires us to find breakthroughs in basic science and build large scale ultra-stable and low energy communication networks with new solutions to further promote the development of internet of things(IoT).Protograph low density parity check(P-LDPC)codes have been widely used in the communication field because of their better error correction performance,lower complexity,and efficient coding and decoding capabilities.In order to make P-LDPC codes meet the requirements of bit error ratio(BER)performance,hardware resource loss and power consumption of communication systems in different application contexts,further design and optimization are needed.This thesis focuses on the design principles of the joint coding system and lossy source coding system based on P-LDPC codes,and provides some ideas for the research of low density parity check(LDPC)codes for IoT.The details are described as follows.(1)Optimization of connection relations of joint coding system:when using channel codes with different row weights and column weights,it is found that different connection relations have different effects on the performance of the joint coding system.In this thesis,P-LDPC codes with different row weights and column weights are used as source codes,and the impact of the type-1 connection basis matrix BL1 on the performance of the system is analyzed from the perspective of both source code and channel code.Moreover,the search algorithm for BL1 under the constraints is proposed based on the joint protograph extrinsic information transfer(JEXIT)and protograph EXIT(PEXIT)to find the optimal BL1 and improve the system performance.Ultimately,both JPEXIT analysis and experimental simulation results show that the optimal solution obtained by the search algorithm proposed in this thesis has significant performance gain in system performance compared to the previous optimal solution.(2)Overall optimization of the joint coding system:the joint basis matrix BJ of the joint coding system is composed of four member elements the source code basis matrix Bs,the channel code basis matrix Bc,the type-1 connection basis matrix BL1 and the type2 connection basis matrix BL2,each of which has a non-negligible impact on the system performance.Unlike previous design and optimization efforts that only target some of the member elements,this thesis proposes a joint design and optimization algorithm based on the idea of Multi-Objective Differential Evolution(MODE).Specifically,the proposed joint design and optimization algorithm is used to design and optimize three member elements Bs,Bc and BL1,and all four member elements Bs,Bc,BL1 and BL2,respectively.The designed basis matrix has a smaller size and element values,which effectively reduces the search complexity and advances the encoding/decoding process.Moreover,the JPEXIT analysis and experimental simulation results show that the scheme obtained in this thesis results in better BER performance of the system compared with the previous optimal scheme.In addition,the system performance is far superior compared to the conventional separated cascade system.(3)Optimization of lossy source coding system:In this thesis,a lossy source coding system based on P-LDPC codes is proposed for binary symmetric sources.At the coding end,the quantization part adopts the Reinforced Belief Propagation(RBP)algorithm;at the decoding end,the linear decoding is used.In addition,according to the factors affecting the performance of P-LDPC codes,two methods of designing P-LDPC codes,DPC-LSC algorithm and NDPC-LSC algorithm,are proposed for this lossy source coding system to improve the rate-distortion performance of the system.Finally,the effects of the parametersγ,L in the RBP algorithm and b-reduction are analyzed by experimental simulations,and the rate-distortion performance of the system is compared after the parameters are determined.The simulation results show that the performance of the P-LDPC codes redesigned by the two methods is significantly better than that of the undesigned P-LDPC codes,and both obtain rate-distortion performance close to the rate-distortion theoretical curve.Moreover,the NDPC-LSC algorithm is used to design and optimize P-LDPC codes from the perspective of balancing the system distortion and the number of iterations,which not only improves the rate-distortion performance of the system but also has more advantages in the number of iterations compared with the DPC-LSC algorithm.