Context-Aware Beam Tracking for 5G mmWave V2I Communications

Vehicles’ mobility causes frequent beam misalignments in millimeter wave (mmWave) vehicle-to-infrastructure (V2I) communications. In 5G systems, beam sweeping is done repeatedly to track a vehicle's mobility and maintain high-quality beam selection. Despite extensive studies on mmWave communications, there are still insufficient discussions on when to trigger beam sweeping for beam tracking. Triggering beam sweeping at an inappropriate time can either incur unnecessary overhead or degrade communication performance due to outdated beam selection. Based on this observation, we explore the problem of making beam-sweeping decisions for 5G mmWave vehicular communications. Considering the regularity in vehicle mobility, we propose a
C
ontext-
a
ware standa
r
d-compati
B
le b
e
am upd
a
te sche
m
e (
CarBeam
) to help the base station determine when to trigger beam sweeping by exploiting the noisy and quantized beam-specific layer-1 reference signal received power feedback from the vehicle. Unlike prior work,
CarBeam
only exploits the procedures and signaling supported in the current 5G beam management framework. Moreover, it can adapt its beam-sweeping decisions to vehicles's mobility for efficient beam tracking. The effectiveness of
CarBeam
is evaluated and demonstrated using the vehicle traces from the TAPASCologne project.