UWB-Based Vehicle Localization using TDoA & AoA

This MATLAB project implements and compares multiple 2D localization algorithms using Ultra-Wideband (UWB) signals — specifically Time Difference of Arrival (TDoA) and Angle of Arrival (AoA) — to estimate a vehicle’s position in a racetrack environment.

---

Overview

• Setup: 10 UWB access points (APs), 4 vehicle-mounted tags, GPS RTK ground truth
• Sampling Rate: 10 Hz
• Tracks: 3 (Obstacle, Straight, Noisy)
• Algorithms:
  • TDoA-only with Hampel & Kalman filtering
  • AoA-only with EKF & NIS-based tuning
  • TDoA + AoA fusion with fallback to AoA
• Performance Metrics: MAE, RMSE, Q95, Availability, Reliability

---

Project Files

Filename	Description
visualizeRawDataStandalone.m	3D AoA ray animation, static AoA visualization, raw TDoA/AoA plots
antennaGT_AoA_processing.m	Raw antenna positions, global AoA elevation/azimuth angles
tdoa_only_multitrack.m	TDoA-only localization + Hampel + Kalman filter (3 tracks)
aoa_ekf_multitrack_final.m	Final AoA-only EKF with NIS-based R tuning and full metrics
tdoa_aoa_fusion_track1.m	TDoA+AoA fusion (Track 1), with fallback and NIS
tdoa_aoa_fusion_track2.m	TDoA+AoA fusion (Track 2), full metrics and plots
tdoa_aoa_fusion_track3.m	TDoA+AoA fusion (Track 3), Kalman filtering, NIS
LNSM_Project_Data.mat	Full dataset: raw TDoA/AoA, AP positions, ground truth
localization_Project_presentation 1.pptx	Final project presentation slides

---

How to Run

1. Open MATLAB (R2021a or later recommended).
2. Add all .m files and LNSM_Project_Data.mat to your working directory.
3. Run any of the analysis scripts directly:

run('aoa_ekf_multitrack_final.m')        % AOA-only EKF
run('tdoa_only_multitrack.m')            % TDoA-only all tracks
run('tdoa_aoa_fusion_track2.m')          % Fusion method on Track 2
run('visualizeRawDataStandalone.m')      % Raw AoA rays and animations