MIT BWSI UAV-SAR System

Developed an airborne Synthetic Aperture Radar (SAR) system capable of imaging ground targets using a UAV. This project integrated signal processing algorithms with real-time flight data.

Technical Approach

• Back-Projection Algorithm: Implemented a “brute force” back-projection algorithm to solve for radar reflectivity ($\sigma(\vec{x})$). This involved integrating pulse data along the flight path to reconstruct the image of a target (a metal can) on the ground.
• State Estimation: Utilized Kalman filtering on data from an external Motion Capture (Mocap) system to precisely estimate the UAV’s position. Accurate position history was critical for the coherent integration of radar signal overlap.
• System Integration: Established a TCP communication protocol to stream telemetry and radar data between the drone and the ground station for processing.