Radar-Assisted Predictive Beamforming for Vehicular Links: Communication Served by Sensing

Abstract

In vehicular networks of the future, sensing and communication functionalities will be intertwined. In this paper, we investigate a radar-assisted predictive beamforming design for vehicle-to-infrastructure (V2I) communication by exploiting the dual-functional radar-communication (DFRC) technique. Aiming for realizing joint sensing and communication functionalities at road side units (RSUs), we present a novel extended Kalman filtering (EKF) framework to track and predict kinematic parameters of each vehicle. By exploiting the radar functionality of the RSU we show that the communication beam tracking overheads can be drastically reduced. To improve the sensing accuracy while guaranteeing the downlink communication sum-rate, we further propose a power allocation scheme for multiple vehicles. Numerical results have shown that the proposed DFRC based beam tracking approach significantly outperforms the communicationonly feedback based technique in the tracking performance. Furthermore, the designed power allocation method is able to achieve a favorable performance trade-off between sensing and communication.

Adaptive and Fast Combined Waveform-Beamforming Design for MMWave Automotive Joint Communication-Radar

“…We show that y m (ρ, δ) in ( 13) can be interpreted as a 2D-DFT measurement of another sparse matrix called the masked Doppler-angle matrix. To define this matrix, we first simplify (13). We define a diagonal matrix containing the scaled DFT of f t (δ) on its diagonal as…”

Distributed Radar-aided Vehicle-to-Vehicle Communication

“…The use of radar signals for improving communication has been studied in [17]- [22]. In [17], a BS equipped with a radar and communication sub-system (operating at different frequencies) is considered, where it is verified that the radar and communication channels share the same dominant paths.…”

“…Once the environment is mapped, communication towards a specific user in a certain location is performed based on predicted propagation paths (including reflected paths). Finally, in [22] a scenario similar to [17] is considered, but with a single joint communication and radar device. No downlink pilots are needed.…”

“…A tracking algorithm predicts the vehicles' states and utilizes these predictions to perform power allocation for both communication and sensing. Among the above works, only [22] considers the unknown data association between radar detections and vehicles. None of the above works develop solution for the V2V scenario.…”

Joint Beamforming Design for Extended Target Estimation and Multiuser Communication

“…To address this issue, the authors of [11], [12] proposed several intrapulse precoding schemes for MIMO DFRC system design, where each communication symbol is represented by a radar snapshot. More recently, these works have been extended to the massive MIMO and Millimeter Wave (mmWave) systems [4], as well as to the scenario of vehicle-to-infrastructure (V2I) links [13].…”