Optimization And Implementation Methods Of The Han-Kobayashi Scheme Using LDPC Codes Over Gaussian Interference Channels

This paper focuses on Gaussian interference channels(GICs) and studies the Han-Kobayashi(HK) coding scheme for the two-user case with the objective of designing implementable channel codes. Specifically, BPSK is employed and low-density parity-check(LDPC) codes are adopted for the use over the channel. Their performance and degree distribution optimization are studied with the following main results and findings:1. Design and implement the LDPC codes codec system under binary input addictive white Gaussian noise channels for the single user case. A linear programming model is proposed to solve the degree distribution optimization problem and some good codes are designed. Performance of the designed codes is studied for finite block lengths through simulations of specific codes picked with the optimized degree distributions, where, for one instance, the results are compared with those of some published codes. Surprisingly, the former is better in the high SNR region;2. Design and implement the LDPC codes codec system under GICs for the case of weak interference. Compared with the single user system, rate splitting and superposition coding techniques are adopted in the transmitter, and the joint decoding algorithm is designed to solve the iterative joint decoding of three decoding components with the help of the state node in the receiver for the two-user system;3. The optimal power allocation algorithm is discussed for the two-user system to maximize the sum rate which is close to the HK achievable rate region. Stability conditions are introduced for GICs, and two-user non-convex programming model is constructed based on the single user linear programming model to solve the degree distribution optimization problem, whose suboptimal solution can still figure out good codes.