Platform
Next-Gen IMOM
IMOM MRA is a multi-resolution, GPU-accelerated electronic warfare analysis prototype designed for ultra-fast processing and visualization of complex radar and jamming environments.
IMOM MRA Technical Overview
View the detailed architecture and operational workflow in the technical documentation.
Overview
Developed as a proof-of-concept using advanced AI coding tools, IMOM MRA is a high-performance Windows-based C++ application designed to test the boundaries of modern system reverse-engineering. It replicates and enhances the primary functions of highly complex GOTS software like IMOM Planner, featuring a modular architecture with both CPU and GPU (CUDA) computation pipelines that allow operators to balance fidelity against compute time.
Key Capabilities
- Accelerated Architecture: Utilizes a highly optimized C++20 codebase with optional CUDA GPU-acceleration for ultra-fast processing and execution.
- Advanced Target Modeling: Calculates radar signal power, receiver noise, signal-to-noise ratios (SNR), and jamming-to-signal metrics (J/S) with computational parity across CPU and GPU backends.
- Multi-Resolution Control: Enables operators to manually adjust grid resolution, angular increments, and single/double precision math pipelines to fine-tune analysis fidelity based on mission needs.
- Terrain & Geodesy Integration: Native DTED terrain ingest via GDAL and precise geodesic calculations using GeographicLib support high-fidelity terrain masking and line-of-sight analysis.
- Geospatial Output: Automatically generates quantitative output artifacts such as tabular CSR/Parquet records and native OGC-compliant KMLs for direct geospatial visualization of vulnerability zones and terrain shadows.
System Architecture
IMOM MRA executes locally as a Windows PC application. The modular system decomposes the workload into distinct components (mra_input, mra_geo, mra_terrain, mra_cpu, mra_gpu, mra_output), bridging cutting-edge algorithmic performance with robust tooling from CMake, YAML-CPP, and Arrow/Parquet.