Tracking and Transmission Design in Terahertz V2I Networks

Abstract

This paper designs the vehicle tracking and resource allocation in the terahertz (THz) vehicle-to-infrastructure communications (V2I) networks, where roadside units (RSUs) equipped with leaky-wave antennas help to estimate the driving states of multiple vehicles and optimize the transmit power and bandwidth per vehicle after receiving the vehicles' feedback. Different from the conventional phased arrays, the leaky-wave antenna has the potential of improving the sensing accuracy with lower system overhead thanks to its unique spatial-spectral coupling feature. The generalized mobile scenario is studied in which vehicles drive at time-varying speeds. A novel unscented Kalman filter (UKF) based solution is proposed to track the vehicles without requirement of addressing the Doppler effect. Based on the estimated states of multiple vehicles, a low-complexity resource allocation method is developed to maximize the sum rate under user fairness concern. Simulation results confirm that the proposed tracking solution can evaluate the propagation angle, vehicle' states and inter-vehicle distance accurately, and the tailored resource allocation method strikes a delicate balance between the sum rate and user fairness in the multi-vehicle V2I scenario.