Analysis of spatial scheduling in downlink vehicular communications: Sub-6 GHz vs mmWave

Abstract

Vehicular communications are gaining a lot of attention for the delivery of enhanced mobility services that require multi-Gbps and low latency connections. In this paper, we focus on Infrastructure-to-Vehicle (I2V) communications where a gNB has to assign spatial resources to a number of connected vehicle users. To efficiently manage the scheduling, we compare the Zero Forcing (ZF) and Maximum Ratio (MR) precoding strategies by evaluating the effect of shifting from sub-6 GHz to millimeter wave (mmWave) frequencies in urban and highway mobility scenarios. We analyze the impact of the geometry of the environment and propagation characteristics at different frequencies in terms of number of users that can be served and spectral efficiency. To model the I2V channel, we integrate realistic traffic conditions generated by SUMO into an accurate channel model based on ray tracing software by WirelessInsite. By numerical results we demonstrate the degradation at mmWave compared to sub-6 GHz on the multiplexing gain. We show the higher efficiency of ZF compared to MR as the former is not limited by inter-user interference, especially in urban scenarios where the number of distinctive eigendirections in space is limited. On the other hand, highway mobility has a more uniform distribution of vehicles that can be conveniently explored by the ZF scheduling to serve more users. Lastly, we show the benefits of adopting a higher number of transmit antennas at mmWave jointly with efficient scheduling to achieve higher spectral efficiency.