Design of capacity-approaching protograph-based LDPC coding systems

In this dissertation, a systematic framework is proposed to design practical protograph-based low-density parity check (LDPC) coding schemes that address simultaneously several important issues: structured coding that permits easy design, low encoding complexity, embedded structure for convenient adaptation to various channel conditions, and performance close to capacity with a reasonable block length. This dissertation consists of four closely inter-related parts. In the first part, the design of rate-compatible protograph codes for the hybrid automatic repeat request protocol is presented. A high-performance family of protograph codes that has the iterative decoding threshold within a gap of a fraction of dB to capacity in the AWGN channel over a wide range of rates is reported. In the second part, protograph-based LDPC coding schemes are designed for half-duplex relay channels. A simple new methodology for evaluating the end-to-end error performance of relay coding systems is then developed and used to highlight the performance of the proposed codes. In the third part, a general mapping method is devised for using protograph-based LDPC codes in bit-interleaved coded modulation. The reported coding scheme operates close to the coded modulation capacity. In the fourth part, the design of rate-compatible protograph codes in inter-symbol interference channels is proposed. The design problem is non-trivial due to the joint design of structured LDPC codes and the state structure of ISI channels using the BCJR equalizer. High-performance protograph-based LDPC codes that have iterative thresholds close to i.u.d capacity of ISI channels are reported.