A New Construction of High Performance LDPC Matrices for Mobile Networks

Area of Science:

• Information theory
• Coding theory
• Network security

Background:

• Secure and reliable information flow is a critical challenge in social Internet of Things (IoT) and mobile networks.
• Ensuring information flow and data integrity remains an open research problem in these domains.

Purpose of the Study:

• To develop novel methods for constructing systematic and regular Low-Density Parity-Check (LDPC) matrices.
• To improve the efficiency and performance of LDPC codes in resource-constrained environments like IoT and mobile networks.

Main Methods:

• Construction of LDPC matrices inspired by the Sarrus method and geometric designs.
• Incorporation of cyclic structures for computational and memory efficiency.
• An optimal post-processing method for deleting girth-four cycles.

Main Results:

• The proposed LDPC codes demonstrate strong performance over additive white Gaussian noise (AWGN) channels when decoded using sum-product LDPC iterative algorithms.
• The developed methods significantly reduce memory consumption.
• Improved convergence speed of the decoder was observed.

Conclusions:

• The novel LDPC matrix construction methods offer an efficient solution for secure and reliable data transmission in social IoT and mobile networks.
• These codes are particularly beneficial for applications requiring low memory footprint and fast decoding.
• The research addresses key challenges in data integrity and information flow for next-generation networks.