An Improved Algorithm for Adaptive Beamforming in MIMO Radar Based on Machine Learning

Adaptive beamforming is a fundamental technique in digital radar signal processing that suppresses interference while enhancing desired signal components.However, the conventional Least Mean Square (LMS) algorithm has inherent limitations in dynamic Multiple Input Multiple Output (MIMO) radar environments, which degrade the balance between convergence speed and steady state stability.To overcome these issues, we introduce a new method called MomentumLMD, which enhances the standard LMS by integrating Momentum and weight Decay.The Momentum component enhances the initial convergence speed, whereas the weight decay mechanism controls the excess kinetic energy to prevent overshoots and oscillations.The simulation results demonstrate that the proposed Least Mean Square with Momentum and Decay algorithm (LMD) outperforms the traditional LMS and standard Momentum + LMS, achieving the lowest Mean Square Error MSE in the steady state and maintaining robustness across a significantly wider hyperparameter range.Therefore, the LMD algorithm is established as the optimal choice for high precision, real-time Multiple Input Multiple Output (MIMO) radar applications.