proposes a Verilog implementation of the Normalized Least Mean Square (NLMS) adaptive algorithm, having a variable step size. The envisaged application is the identification of an unknown system. First the convergence of derived LMS algorithm

. The normalized least mean square is the most popular due to its simplicity. The stability of the basic NLMS is controlled by a fixed step size. This parameter also governs the rate of convergence, speed of tracking ability and the amount of steady-