Previous Abstract Return to Session F3 Next Abstract

Session F3: Navigating in Challenged Environments (e.g., Urban, Indoor and Subsurface Navigation)

RAD-NAV: SAR Datasets for RF-Aided Navigation in GPS-Denied Environments
Andrew Willis, UNC Charlotte Pamela Card, Air Force Research Laboratory Kevin Brink, Air Force Research Laboratory
Location: Room 6-8
Date/Time: Monday, Jun. 1, 2:10 p.m.

This paper presents the development of two airborne radar pod systems and associated datasets designed explicitly to accelerate the development of synthetic aperture radar (SAR) based algorithms for navigation in GPS-denied and challenged environments. Although coherent radar systems are widely used for imaging applications, progress in leveraging SAR observables for navigational aiding has been constrained by the scarcity of accessible, well-documented raw SAR datasets and by the complexity of software pipelines required to access and interpret raw complex I/Q data produced by modern radar hardware.
The radar pods described in this work are instrumented to support coherent transmit–receive operation and high-fidelity motion sensing, enabling the direct study of SAR and interferometric SAR (iSAR) measurements as navigation-relevant signals. One pod employs a four-antenna architecture suitable for SAR and InSAR image formation and interferometric phase measurements, while the second pod is optimized for stripmap SAR collection using unusually high pulse repetition frequencies (PRFs), up to 20 kHz, enabling dense temporal sampling of radar returns. These characteristics make the systems well suited for investigating the relationship between RF backscatter, platform motion, and navigational state estimation in the absence of GNSS.
To address barriers to algorithm development, this work introduces curated radar datasets and documentation that expose raw SAR I/Q data alongside progressively processed products. Dataset organization follows conventions used by established SAR repositories, including NASA’s Alaska Satellite Facility and ESA’s Sentinel-1, with products categorized as Level 0 raw data and Level 1+ processed SAR outputs. The paper presents a system-level measurement and motion model that precisely characterizes how recorded RF signals are formed within the radar hardware and how these measurements map to navigational parameters such as position, velocity, and altitude.
Representative data examples and initial alternative navigation (alt-Nav) demonstrations are presented, illustrating how SAR observables can be exploited for RF-aided navigation in GPS-denied contexts. By lowering data access barriers and providing rigorously documented SAR measurements, this work aims to reduce development friction and accelerate research into resilient navigation methods based on coherent radar sensing.

Previous Abstract Return to Session F3 Next Abstract