-
Notifications
You must be signed in to change notification settings - Fork 1
Introduction and Goals
(Learn more about the FERS project and the current MSc research objectives)
FERS (Flexible Extensible Radar Simulator) is a powerful, open-source C++ based simulation tool designed for modeling a wide variety of traditional and modern radar systems. Developed initially by Marc Brooker and Michael Inggs at the Radar Research and Signal Group (RRSG), FERS allows for detailed signal-level simulations and comprehensive performance assessments of radar systems under diverse operational conditions and configurations.
The simulator has recently undergone a significant modernization effort, migrating the codebase from C++98/03 to modern C++20/23 standards. This modernization has drastically improved performance, maintainability, and modularity, resulting in:
- Reduced cyclomatic complexity (from 1-15 to 1-10).
- A significant decrease in code quality warnings (from 3243 to 6).
- An improved technical debt rating (from C to A).
- The introduction of a comprehensive regression testing suite with ~90% line coverage.
These enhancements have laid a solid foundation for future development and expansion.
- Signal-Level Simulation: Detailed modeling of radar signal returns, including Doppler shift and phase evolution.
- System Versatility: Support for Monostatic and Multistatic radar configurations.
- Radar Modes: Simulation of Continuous Wave (CW) and Pulsed radar systems. (FMCW is a current development priority).
- Propagation Effects: Modeling of multipath propagation effects (currently limited to a single flat surface).
- Data Export: Comprehensive output options including CSV, KML, and HDF5 formats for analysis.
- Modern Architecture: Leverages smart pointers for enhanced memory management and a global thread pool for efficient multithreading.
This Master's research project aims to build upon the recent modernization by significantly enhancing the usability, reliability, and extensibility of FERS. The overarching goal is to transform FERS into a more robust, user-friendly, and versatile tool for both seasoned radar researchers and newcomers to the field.
The key objectives for this MSc research have been refined to focus on two high-impact areas: delivering a best-in-class user experience for scenario creation, and expanding the core simulation engine's capabilities and physical accuracy.
-
Revolutionize Usability with the FERS World Builder UI:
- Objective: Develop an intuitive, standalone desktop application for creating, configuring, and visualizing FERS simulation scenarios. The goal is to eliminate the steep learning curve associated with manually writing complex XML files.
- Technology: The UI will be built using a modern web-based stack (React/CesiumJS/Electron) to provide a responsive user interface and a high-fidelity, interactive 3D globe for geolocated scenario building.
- Core Functionality: The UI will act as a powerful "World Builder," enabling users to visually place and configure platforms, manage motion paths, and receive immediate visual feedback on their scenario setup. It will handle the full import and export of FERS XML configuration files and will be able to run the FERS backend directly.
-
Enhance Simulation Engine Reliability and Capabilities:
- FMCW Radar Mode Implementation: Implement robust, end-to-end support for Frequency-Modulated Continuous Wave (FMCW) radar simulations. This is the top-priority feature addition, requiring the development of an internal de-chirping model to ensure accurate signal-level simulation.
-
Address Core Architectural Limitations: A significant part of the research involves moving FERS beyond its academic origins by refactoring critical parts of the simulation engine. This includes:
- Fixing the physically inaccurate waveform transformations in moving transmitter scenarios.
- Redesigning flawed oscillator offset models to correctly simulate multi-static systems.
- Replacing the inefficient "long pulse" hack for CW mode with a native implementation.
- Validating and improving the reliability of the core noise models and receiver signal chain.
-
Improve Project Accessibility and Maintainability:
- Integrated KML Visualization: The standalone KML generation tool has been refactored and fully integrated into the FERS backend, providing a streamlined option for visualizing scenarios in external tools like Google Earth.
- Comprehensive Validation: Develop and expand robust unit and scenario-based test cases to rigorously validate all new features and core model enhancements against theoretical predictions.
- Streamline Codebase: Fully remove the deprecated legacy Python integration to simplify the build process and eliminate maintenance overhead.
By achieving these focused objectives, this research will produce an enhanced version of FERS that is not only more powerful and physically accurate but also significantly easier to learn, use, and extend, fostering its adoption within the RRSG and the broader radar community.