Divergence Based Optimization of Multi Rate QC-LDPC Code Design

QC-LDPC codes need large amount of iterations to converge, decoding delay in these codes is still quite difficult. This paper presents a divergence-based optimization technique for QC-LDPC codes, in which bit-level reliability is characterized using Kullback-Leibler (KL) divergence. An adaptive puncturing and bit-selection strategy is proposed, which selects highly reliable nodes during the iterative decoding process. To improve reliability, a Hybrid Automatic Repeat Request with Incremental Redundancy (HARQ-IR) mechanism is incorporated, allowing the accumulation of soft information across retransmissions. The proposed approach leverages the observed relationship between divergence and bit reliability to accelerate decoder convergence and reduce computational complexity. Simulation outcomes in an additive white Gaussian noise (AWGN) channel indicate that the suggested strategy attains bit error rate (BER) performance that is either equivalent to or enhanced. This approach can attain BER levels ranging from 10^-6 to 10^-7, while at the same time decreasing the number of decoding iterations by 40% compared to conventional methods. The proposed method is appropriate for next- generation communication systems as it achieves an equilibrium between performance and decoding complexity.