Abstract

Synthetic Aperture Radar (SAR) is a powerful tool for imaging and reconnaissance, typically operated in a monostatic configuration. Bistatic geometries spatially separate the transmitter and receiver, capturing unique target scattering information, while passive methods leverage existing environmental signals of opportunity. With the advent of massive Low Earth Orbit (LEO) broadband constellations like SpaceX’s Starlink, high-bandwidth reference signals with large Doppler modulation now exist globally, offering bistatic SAR imaging potential worldwide. This research presents an end-to-end simulation to demonstrate the passive exploitation of Starlink communication waveforms for tomographic bistatic SAR imaging. Utilizing high-fidelity Computer-Aided Design (CAD) models, the simulation generated a synthetic dataset of over 19,000 images. This dataset was then used to train a transfer learning model, demonstrating the feasibility of target characterization even with partially resolved imagery. Finally, post-processing of real-world Starlink signals validates the ability to passively recover the reference channel information required for practical image formation.

Publication Date

7-28-2026

Document Type

Dissertation

Student Type

Graduate

Degree Name

Imaging Science (Ph.D.)

Department, Program, or Center

Chester F. Carlson Center for Imaging Science

College

College of Science

Advisor

Charles Bachmann

Advisor/Committee Member

James Albano

Advisor/Committee Member

Alireza Vahid

Comments

This thesis has been embargoed. The full-text will be available on or around 8/9/2027.

Campus

RIT – Main Campus

Download

Share

COinS