Award Date

5-15-2026

Degree Type

Thesis

Degree Name

Master of Science in Engineering (MSE)

Department

Electrical and Computer Engineering

First Committee Member

Peter Stubberud

Second Committee Member

Ebrahim Saberinia

Third Committee Member

Sahjendra Singh

Fourth Committee Member

Brendan O'Toole

Number of Pages

210

Abstract

Analog frequency sampling filters (FSFs) provide an efficient means of realizing finite impulse response (FIR)-like behavior in continuous-time systems, but their practical implementation is constrained by the requirement for perfect pole-zero cancellation along the imaginary axis. Because exact cancellation is physically unattainable due to component variations, ideal linear-phase Type 1 analog FSFs exhibit uncancelled poles that result in system instability. To address this limitation, this thesis introduces a near-linear-phase design framework for Type 1 analog FSFs that achieves both stability and design flexibility through the inclusion of a damping constant, ρ, which shifts the poles into the left half of the complex plane. The proposed method formulates the filter design as an unconstrained optimization problem and uses a trust-region optimization technique to minimize a weighted cost function that combines passband fidelity, stopband attenuation, and phase deviation. Unlike linear-phase FSFs, which restrict optimization freedom under the linear-phase constraint, the near-linear-phase FSF provides additional degrees of freedom that reduce interpolation error while maintaining near-linear phase characteristics. By adjusting the weighting parameters, the designer can control trade-offs between amplitude fidelity, phase linearity, and interpolation accuracy to achieve a desired frequency response. Compared to traditional FSF designs, the proposed framework enables stable, realizable analog implementations without requiring ideal pole-zero cancellation. The near-linear-phase analog FSF is designed and simulated in MATLAB to match a desired frequency response, and its performance is compared with an ideal linear-phase FSF to evaluate reductions in interpolation error between the passband and stopband regions. Both filter designs are subsequently implemented and simulated in LTspice, and the resulting frequency responses are compared with the MATLAB simulations. The close agreement between the LT-spice and MATLAB results validates the accuracy of the proposed near-linear-phase analog FSF and demonstrates its practical viability for stable, high-speed, and narrow-band signal processing applications.

Keywords

Analog Filters; Cost Function; Frequency Sampling; Hessian; Near-Linear Phase; Trust Region Optimization

Disciplines

Electrical and Computer Engineering | Electrical and Electronics | Engineering

File Format

PDF

File Size

8300 KB

Degree Grantor

University of Nevada, Las Vegas

Language

English

Rights

IN COPYRIGHT. For more information about this rights statement, please visit http://rightsstatements.org/vocab/InC/1.0/

Download


Share

COinS