The Research Of Encoding And Decoding For LDPC Codes And Its Application In Telemetry System

Low-Density Parity-Check (LDPC) codes whose performance is very close to the Shannon limit are a class of linear block codes. With strong error control strength, lower decoding complexity, lower error floor, and many others advantages, the prospects of LDPC codes in the reliable transmission of information has attracted the academic and the industrial field. Hence it is now a research hotspot in the coding field.This dissertation, based on a scientific research project, and the method of the theoretical analysis with computer simulations, studies the principle of decoding LDPC codes, parity check matrix of the structure and the decoding algorithm. And then it builds simulation platform for PCM/FM, applying the LDPC codes to the telemetry system. At last it gives the simulation analysia. The major work done in the following areas:Firstly, randomly constructed LDPC codes always have high complexity, and taking much storage resources, and difficultity in the realization of the hardware. To overcome this shortcoming, a low coding complexity of quasi-cyclic LDPC code construction method has been proposed. It reduced the demand of coding delay and encoding hardware resources.Secondly, to improve the performance and reduce computational complexity of LDPC codes, a differential decoding algorithm based on min-sum is introduced. In the algorithm, the messages passed in the bipartite graph of LDPC codes are the probability differential values, and the updates of check nodes adopt min-sum algorithm. Simulation results show that the min-sum based on differential decoding algorithm has 0.4dB SNRs gain than the UMP BP-Based decoding algorithm at a bit error rate of 10-5 in an additive white Gaussian noise (AWGN) channel with the same decoding complexity.Thirdly, according to no-convergence of falsely oscillating iteration in the LDPC decoding at the range from medium to high SNRs, we proposed a modified LDPC belief propagation (BP) decoding algorithm, i.e. soft value zero-forcing BP algorithm. By setting extrinsic information of the oscillating iteration bit nodes into zero, the impacts on the iteratively decoding from false channel information is greatly reduced. And it also improves the performance of the decoding algorithm. Furthermore, a decision criterion of oscillating iteration nodes is presented to increase the accuracy of the decision. Simulation results show that the proposed algorithm has better decoding performance than that of the BP algorithm with the same iterations at medium and high SNRs. It has 0.4dB performance gain better than the UMP BP-Based decoding algorithm at a bit error rate of10-5. Finally, based on the characteristics of the PCM/FM telemetry system, the architecture model of the LDPC encoded PCM/FM telemetry system was built. In order to measure the performance of the system, The QC-LDPC codes constructed in this paper have been applied in the system and the system performance was analyzed. The simulation results show that the PCM/FM telemetry system with LDPC encoding can get a coding gain of about 8 dB at a bit error rate of11-5. Compared with mature Turbo codes, LDPC codes have more than 2～3 dB SNRs gain, lower error floor and complexity.