The article proposes a method of assessing information transmission reliability by using the output normalized logarithmic ratio of the likelihood function (LRLF) of the decoder. Based on the evaluation, the method allows adapting system parameters with turbo codes (TC) or LDPC code. This method can be used in combination with other methods of parametric and structural adaptation using turbo codes or LDPC codes.

Low-Density Parity-Check (LDPC) codes are among the most effective modern error-correcting codes due to their excellent correction performance and highly parallel decoding scheme. Moreover, the nonbinary extension of such codes further increases performance in the short-block regime. In this paper, we review the key elements for the construction of implementation-oriented binary and nonbinary codes. These Quasi-Cyclic LDPC (QC-LDPC) codes additionally feature efficient encoder and decoder implementation frameworks. We then present a versatile algorithm for the construction of both binary and nonbinary QC-LDPC codes that have low encoding complexity and an optimized corresponding graph structure. Our algorithm uses a progressive edge growth algorithm, modified for QC-LDPC graph construction, and then performs an iterative global search for optimized cyclic shift values within the QC-LDPC circulants. Strong error correction performance is achieved by minimizing the number of short cycles, and cycles with low external connectivity, within the code graph. We validate this approach via error rate simulations of a transmission system model featuring an LDPC coder-decoder, digital modulation, and additive white Gaussian noise channels. The obtained numerical results validate the effectiveness of the proposed construction algorithm, with a number of constructed codes exhibiting either similar or superior performance to industry standard binary codes and selected nonbinary codes from the literature.

In this paper, the performance of Low-Density Parity-Check (LDPC) codes is improved, which leads to reduce the complexity of hard-decision Bit-Flipping (BF) decoding by utilizing the Artificial Spider Algorithm (ASA). The ASA is used to solve the optimization problem of decoding thresholds. Two decoding thresholds are used to flip multiple bits in each round of iteration to reduce the probability of errors and accelerate decoding convergence speed while improving decoding performance. These errors occur every time the bits are flipped. Then, the BF algorithm with a low-complexity optimizer only requires real number operations before iteration and logical operations in each iteration. The ASA is better than the optimized decoding scheme that uses the Particle Swarm Optimization (PSO) algorithm. The proposed scheme can improve the performance of wireless network applications with good proficiency and results. Simulation results show that the ASAbased algorithm for solving highly nonlinear unconstrained problems exhibits fast decoding convergence speed and excellent decoding performance. Thus, it is suitable for applications in broadband wireless networks.