A novel and efficient Hybrid Least Mean Square (HLMS) adaptive algorithm for system identification

System identification is one of the most interesting applications for adaptive filters, especially for the LMS algorithm due to its robustness and computational simplicity. Owing to the limitations of LMS algorithm in terms of convergence and non-linearity, many variants of LMS has been tried for various applications. In this paper, we propose an effective approach to identify unknown system adaptively by using Hybrid Least Mean Square (HLMS), which can be considered as yet another variant of standard LMS algorithm. It uses a variable convergence factor of NLMS for faster convergence and quantization of error signal to reduce mean square error (MSE) and complexity. The leakage factor of Leaky LMS is incorporated in the weight update equation so that the algorithm works efficiently even in noisy environment. This gives robust performance even at larger step size where the traditional LMS tends to become unstable. Simulation results show that the proposed algorithm is superior to LMS and its other variants in terms of convergence rate, robustness and mean square error (MSE).