GARGOYLE: Generative AI RF Geospatial Overlays for Localization of Emitters

In urban areas, localizing Radio Frequency (RF) emitting sources is complicated by multipath propagation, where signals bounce off structures and create multiple paths between emitter and receiver. This interferes with localization algorithms and TTPs that assume a single, direct signal path. GARGOYLE uses applied AI to characterize available GIS data, it then leverages this understanding along with near-real-time on-target signal sensors to apply a combination of methods (e.g., ray tracing, particle-based simulation) to simulate and predict RF propagation for candidate emitters as a mechanism to facilitate improved electronic warfare (EW) mission planning and execution. GARGOYLE intends to present this information through both 2D tablet and 3D Augmented Reality (AR) sandtable visualizations (validated at a recent AFSOC-supported T&E event) to provide Operators with actionable intel. GARGOYLE seeks to provide SOF EW Operators with a tool for near-real time 3D RF signal propagation modeling that uses generative AI to contextualize models to dense urban environments (via GIS data) to facilitate more rapid and accurate localization of RF emitting devices (e.g., a targetĺs smartphone) during direction finding (DFĺing) or similar threat detection & characterization TTPs. This will both enhance Operator survivability (by enabling them to spend less time in potentially hostile regions and have more refined plans for where they need to run survey and DFĺing TTPs) and increase Operator situational awareness (SA; through better understanding of RF terrain) ľ ultimately improving the likelihood of positive mission outcomes.á á