Complex High-Resolution Imaging Radar Scene Generation for Real-Time Applications

Establishing design viability of sensors integrating guidance, navigation and control in a ground test environment is a critical step before fielding of new or updating of existing vehicles. This part of the development and sustainment process requires continual hardware-in-the-loop (HWIL) testing. In order to enable testing of cutting-edge sensors, continuous development of test technologies is needed and pose a high demand on scene phenomenology modeling and target simulators, including real-time RF target simulations for imaging radars. Example scene update rates are at around 1200 Hz, with resolution requirements of 1 million scatterers per target scene. High-fidelity ray-tracing, including physical scattering on object surfaces will be required for realistic, high-resolution radar scene generation, meeting current sensor capabilities and requirements. Consequently, we propose to build upon an existing real-time scene generator validated for the EO/IR frequency bands and plan to develop plug-in channels for the RF frequency bands. Waveform generation will be directly implemented into the scene generator. Real-time ray-tracing is achieved by new GPU graphics cards in combination with new algorithmic methods, able to cast billions of rays per second, and rendering billions of triangles in only milliseconds. Besides CPUs and GPUs, we will explore FPGAs, RFSOCs and DSPs and their impact on acceleration potential. This tool will thus be able to aid in the analysis and design process of high-speed sensors and vehicles and as such, due to its cost-effective, efficient and high-quality solution approach, will play a fundamental role and provide significant value for its outlined application with high transition potential to DoD and further transition and commercialization to prime contractors and private industry.