Development and Simulation of Leaky LMS Beamforming Algorithm

The LMS algorithm is one of the important adaptive filtering algorithms, which are applicable to the adaptive calculation and good stability. However, as the LMS algorithm is based on the rough gradient estimation, there is an inherent conflict among the convergence rate and steady-state error. In order to overcome the contradiction, some modified variable step-size LMS algorithms have been proposed in recent years. The VS-LMS algorithm adjusts the step factor by establishing a nonlinear relationship between step factor and error signal. It improves the convergence rate, but steady-state error and ability of tracking is still a little dissatisfied, especially in a low SNIR condition, which restricts its applications. This paper presents a leaky LMS and Novel VS-LMS algorithm, designed to overcome the slow convergence of standard LMS in cases of high input eigenvalue spread.