Power and area efficient FIR filter using Radix- 2^r multiplier for de-noise the electrooculography (EOG) signal

Area of Science:

• Digital Signal Processing
• Biomedical Engineering
• VLSI Design

Background:

• Electrooculography (EOG) signals are crucial for diagnosing eye movement disorders.
• Traditional Finite Impulse Response (FIR) filters for EOG signal denoising often suffer from high computational complexity, leading to increased delay and power consumption.
• Optimizing FIR filter design is essential for efficient real-time EOG signal processing.

Purpose of the Study:

• To introduce a novel FIR filter architecture for enhanced Electrooculography (EOG) signal denoising.
• To reduce the hardware resource utilization, power consumption, and signal processing delay of FIR filters.
• To improve the overall efficiency and performance of EOG signal processing systems.

Main Methods:

• A novel FIR filter architecture is proposed, integrating a Radix-2r multiplier with 4:2 and 3:2 compressors.
• Compressors are utilized to replace traditional ripple-carry adders for partial product summation, optimizing the multiplier's performance.
• The filter design is implemented using gate-level Verilog Hardware Description Language (HDL) and verified using ModelSim and Altera DSP Builder.

Main Results:

• The proposed FIR filter demonstrates significant reductions in hardware resources and power consumption.
• Achieved a 71.48% reduction in area, 73.28% reduction in power, and 51.84% reduction in delay compared to conventional designs.
• The Area-Delay Product (ADP) and Energy-per-operation (EPS) were reduced by 86.22% and 92.38%, respectively, outperforming the SOPOT filter.

Conclusions:

• The novel FIR filter architecture offers substantial improvements in efficiency for EOG signal denoising.
• The use of Radix-2r multipliers and compressors provides a highly optimized solution for low-power, high-speed digital signal processing applications.
• This design presents a promising advancement for real-time biomedical signal processing, particularly for EOG-based systems.