Search

Abstract This article introduces a novel FIR filter using a Radix-2r multiplier combined with 4:2 and 3:2 compressors for denoising Electrooculography (EOG) signals. This approach replaces traditional ripple-carry adders with compressors to add the partial products generated by the Radix-2r multiplier, resulting in reduced delay and energy consumption. The FIR filter is implemented in gate-level Verilog HDL, verified using ModelSim and Altera DSP Builder, and synthesized with the Cadence RTL compiler. Compared to conventional designs, the proposed filter achieves a 71.48% reduction in area, 73.28% reduction in power, 51.84% reduction in delay, 86.22% reduction in the area-delay product (ADP), and 92.38% reduction in energy-per-operation (EPS), significantly outperforming the SOPOT filter.