Leaky IPNLMS Algorithm for Adaptive Equalization in QAM Systems over Underwater Acoustic Multipath Channels

This paper proposes a Leaky Incremental Proportional Normalized Least Mean Squares (Leaky-IPNLMS) algorithm for adaptive equalization in underwater acoustic (UWA) communication systems, specifically for 16-QAM modulation. The algorithm enhances performance by combining the NLMS framework with a dynamic leakage factor to improve convergence speed and stability in multipath channels. Simulations over realistic UWA multipath channels show that the Leaky-IPNLMS algorithm outperforms traditional LMS, NLMS, and VSS-NLMS methods in terms of Symbol Error Rate (SER) and convergence time. The results demonstrate its robustness in high-noise environments, offering a balance between equalizer performance and computational efficiency, making it a promising solution for UWA communication.