Low Power and Area Efficient FIR Filter Using Adaptive LMS Algorithm

The paper presents low power and area efficient LMS (Least Mean Square) adaptive filter. Least Mean Squares algorithm used in many DSP applications to process the signal and designed for generating filter coefficients based on the processed signal. The proposed method reduces the power consumption and area due to adaptation of the weights generated internally according to the corrected output and previous weights. The weights area calculated using LMS algorithm. The proposed simulation is done using Modelsim, structure coded using Verilog HDL and synthesized by Xilinx ISE. The power is calculated using Xilinx power estimator by targeting to FPGA. The observed results for the proposed system improved 46 % in area and 23 % in power compared to conventional FIR filter.