Paper Title: Federated Learning-Based Joint Radar-Communication mmWave Beamtracking for V2X-Communications
Conference Name: The 3rd International Conference on M3 IT Convergence, August 7-9, 2024, Concorde Hotel Kuala Lumpur
Abstract: Robust beamforming in millimeter-wave (mmWave)
communication, vital for Vehicle-to-everything communication,
faces challenges like short-range links due to path loss and
obstacles. Privacy concerns and bandwidth constraints make
transmitting the entire dataset impractical. To tackle this, a
federated learning (FL)-based joint radar and communication
mmWave beamtracking approach, FL-JRC, is proposed. FLJRC
employs a federated convolutional neural network (CNN)
for vehicle users (local clients) and road side user (RSU) (main
server) to estimate arrival/departure angles, optimizing beam
tracking. Simulation results demonstrate superior performance
compared to contemporary methods and two Kalman filter
versions (extended and unscented).