Drone localisation and tracking using a distributed FMCW MIMO radar network
Term
4. semester
Education
Publication year
2025
Submitted on
2025-06-03
Pages
82
Abstract
The proliferation of unmanned aerial vehicles (UAVs) presents both opportunities and threats, necessitating robust detection and tracking systems. This thesis explores the feasibility of deploying the multiple radar Bayesian localisation and tracking (MRBLaT) algorithm on real-world data obtained from distributed, off-the-shelf frequency-modulated continuous-wave (FMCW) multiple-input multiple-output (MIMO) radar systems. A comprehensive signal model for FMCW radar is developed, leading to the development of the Python-based Tracking and Mimo radAR Simulation (TMARS) package. The MRBLaT algorithm is integrated into this package to evaluate performance under both simulated and real-world conditions. Measure- ments are collected using three Texas Instruments AWR1843BOOST radar modules operating in a synchronised, distributed configuration. Experimental results demonstrate the MRBLaT algorithm’s ability to track a drone under real-world conditions with significant accuracy, es- pecially in low signal-to-noise ratio (SNR) scenarios. Challenges related to multipath fading, clutter, synchronisation, and non-convex optimisation are identified as key difficulties of real- world deployment. The findings validate MRBLaT as a promising approach for distributed radar-based drone tracking and lay the groundwork for future improvements.
Keywords
Documents