As a Modeling, Simulation, and Analysis Engineer at Northrop Grumman, I was a core developer on a high-fidelity C++ radar system simulation exceeding one million lines of code. The simulation integrates with the operational flight program (OFP) to evaluate radar performance in diverse scenarios. Running on a high-performance computing cluster of more than 80 nodes, it produces over a petabyte of data monthly, requiring scalable and efficient data analysis techniques.
I designed and implemented a sophisticated electronic attack clutter model to simulate adversarial electronic transmissions affecting radar detection. I also optimized a clutter mesh generation algorithm in C++ using Intel Integrated Performance Primitives (IPP) for SIMD acceleration. These enhancements significantly reduced runtime and enabled higher-fidelity simulations. Performance tuning was supported by profiling with Intel VTune.
I contributed to development workflow improvements by helping set up Git-based version control and CI/CD automation with GitLab, which boosted team efficiency. I also supported infrastructure upgrades, including migrating our HPC environment to a newer Red Hat Enterprise Linux version, resolving dependency issues involving the IPP library, GCC, and CMake, and building internal tools to simplify the transition for users.
In addition to simulation work, I developed tools for analyzing IQ data to extract metrics like angle of arrival, range, and velocity. These tools featured just-in-time parsing and an LRU caching system to manage large datasets efficiently. I also wrote comprehensive technical documentation detailing model architecture and underlying mathematical frameworks. Separately, I built PyCAM, a web application for control account managers to track program costs, using Flask, MongoDB, and custom JavaScript to deliver a responsive and user-friendly interface.
