On Applications Of Low Density Parity-Check Codes In DAMB And VWDK Systems

Recent advances in error correcting codes have shown that irregular low density parity-check (LDPC) codes attracted considerable attentions owing to their capacity- approaching performance and low complexity iterative decoding. To improve the efficiency of transmitting encoded data in wireless channel, in this dissertation, we propose LDPC codes for implementation of the Digital Amplitude Modulation Broadcasting (DAMB) and the Very-minimum Waveform Difference Keying (VWDK) systems.In DAMB systems, achieving good audio quality becomes a challenge due to its limited band-width of 9 or 10 KHz and the very bad fading channels. Therefore, DAMB needs highly efficient channel coding schemes. In this thesis, a series of coded modulation schemes based on irregular LDPC codes are employed, after exploration of international standard proposed by Digital Radio Mondiale (DRM). Firstly, three decoding strategies using soft decision are provided for Multilevel Coding (MLC) schemes based on Rate Compatible Punctured Convolutional (RCPC) codes. And then, Bit-Interleaved Coded Modulation, MLC and iterative demapping and decoding schemes based on irregular LDPC codes are investigated for DAMB systems. Finally, using Extrinsic Information Transfer (EXIT) chart technique, we optimize the irregular LDPC codes for DAMB systems and achieve satisfactory results. Simulation results show that our schemes are more efficient than the RCPC coded DRM system on various broadcast channels. The optimized irregular LDPC codes can outperform Turbo codes, especially in short wave channel.The pivotal thinking of VWDK is to make the difference of two modulation waves minimum, which makes coherent detection more difficult, then results in poor SNR performance. Therefore, powerful channel coding schemes are essential for VWDK modulation. The paper investigates the application of LDPC codes for VWDK modulation. We modify Belief Propagation (BP) decoding algorithm, and prove symmetry on AWGN...