Optimization of LDPC Decoding Using Layered Min-Sum and Early Stopping Techniques for Energy-Efficient and Reliable Wireless Sensor Networks

Abstract

This research investigates the optimization of Low-Density Parity-Check (LDPC) decoding for enhanced energy efficiency and reliability in Wireless Sensor Networks (WSNs), which are inherently constrained by limited power, processing capability, and bandwidth. Recognizing the limitations of conventional error control techniques, particularly the high energy consumption associated with standard LDPC decoding, the study proposes an enhanced decoding model based on the Layered Min-Sum (LMS) algorithm integrated with an Early Stopping mechanism. The model was implemented and evaluated using MATLAB under an Additive White Gaussian Noise (AWGN) channel with Binary Phase Shift Keying (BPSK) modulation across Eb/No values ranging from 1 to 9 dB. Performance metrics considered include Bit Error Rate (BER), Decoding Energy per Bit (DEB), and iteration count. Simulation results demonstrate that the developed LMS with Early Stopping significantly outperforms the conventional Flooding Min-Sum (FMS) decoder, achieving a mean DEB reduction from 1.2960 nJ/bit to 0.7174 nJ/bit (approximately 45% energy savings) and a reduction in average iteration count from 18.40 to 10.23 (about 44% improvement). Additionally, the developed model maintains superior error performance, recording a slightly lower mean BER of 0.1055 compared to 0.10612 for FMS. These results confirm that the integration of layered scheduling and early stopping effectively reduces computational overhead and energy consumption without compromising error correction capability. The study concludes that the optimized LDPC decoding is a viable and efficient solution for improving the performance and longevity of energy-constrained WSNs, thereby supporting more reliable and sustainable wireless communication systems.