Study On The System Of Joint LDPC Codes With Bit-interleaved Coded Modulation Iterative Demodulation

Motivated by its low complexity, strong agility, high spectrum efficiency, andexcellent performance, Low Density Parity Check (LDPC) coded bit-interleave codedmodulation with iterative decoding (BICM-ID) has been a key technique of nextgeneration of wireless communication. And it has been widely used in moderncommunication systems. The current researches of LDPC-BICM-ID system, whichinclude the mapping constellation design, interleaving scheme and iterative decodingalgorithm, have been studied rigorously. However, there are still many problems need tobe resolved, especially for the choice of signal mapping in LDPC-BICM-ID systemwhich does not exist gray mapping and how to improve decoding gain between themodulator and the decoder. This dissertation focus on above two topics, then do furtherinvestigation on BICM-ID system based on LDPC codes.1）This paper do a systematic study on the basic principles of the LDPC codes,encoding method of LDPC codes, decoding method of LDPC codes and the technologyof BICM-ID. The advantages and disadvantages of two LDPC codes encoding methodsare analyzed and compared; Probability of Belief-Propagation (BP) decoding algorithmand the LLR BP decoding algorithm are derived. The structure and working principle ofBICM-ID is researched, and the simulation results indicate that the system with iterativedecoding has better bit error performance than BICM and TCM.2）Bit-interleaved LDPC coded modulation iterative decoding, which combinesLDPC codes and BICM-ID, is researched as an emphasis part, including built thearchitecture of LDPC-BICM-ID system, derive the process of joint demodulation withiterative decoding and give an efficient soft demodulation algorithm. Based on theimpact of signal labeling, verified constellation mapping based on mutual informationcriterion are simulated, and a new signal labeling designed criterion is proposed for bitinterleaved LDPC coded star-16QAM modulation with iteratively decoding in thisdissertation, it is in order to maximize the Euclidean square and minimum decisionregion for this system. A new star-16QAM signal labeling is designed based on it.Comparative study of different labeling is done through analysis and computersimulation, the results under AWGN channel show that the new designed signal labelingachieves a big gain. Also, the influence of other parameters on system performance areresearched, such as code length and column heavy of LDPC codes, interleaving and the number of iterations of LDPC-BICM-ID system.3） Research on the effective method analyzing convergence of iterativedecoding-Extrinsic Information Transfer chart(EXIT). Analyzed the iterative process ofBICM-ID based LDPC codes and the influence of some labeling maps on the systemperformance by EXIT chart in AWGN channel, we validate the conclusion which is gotby simulation. In order to improve the performance of BICM-ID based on codesdramatically by comparing to BICM with the same constituent code. The reason of thisphenomenon is analyzed and an improved joint demodulate decoding method isproposed by combining the characteristics of LDPC and BICM-ID. Simulation resultsshow that the performance of this enhanced LDPC-BICM-ID is improved.