Scheduling & resources sharing technique for adaptive LMS filter

This paper presents a scheduling & resources sharing technique for adaptive LMS (Least-Mean-Square) filter with low power consumption and low resources utilization. It used an 8-bit fixed-point arithmetic representation for the adaptive filter system based on FPGA. The design has been synthesized and simulated in Xilinx ISE. LMS algorithm is widely used in the adaptive filter for noise cancellation system or hearing aid system. We verified the parameter step size u, through simulated the LMS algorithm on MATLAB. And then, we implemented and simulated the original LMS circuit and the proposed LMS circuit which use the scheduling & resources sharing technique design in Xilinx ISE. We found that the proposed circuit uses resources less than the original circuit. Based on the XPower Analyzer tool, the power consumption is estimated. The results showed that the proposed circuit dissipates less power than the original circuit. It nearly economizes 36.4% power consumption compared with the original one.