Intelligent Link Adaptation in 802.11 Vehicular Networks: Challenges and Solutions

By providing effective wireless vehicle-to-everything (V2X) communication, it is expected that the Internet of Vehicles will be provisioned, whereby various automotive service requirements could be satisfied (enhanced road safety, improved transportation efficiency, etc.). Despite the successful standardization of 802.11-based communication protocols in vehicular networks (e.g., 802.11p, 802.11af), there remains an essential issue, which is to adjust the transmission rate and mode to match the highly dynamic channel conditions in vehicular scenarios. In this article, we first investigate the requirements for intelligent link adaptation (LA) in 802.11 V2X, and then present a data-driven learning-based LA architecture that fits the road channel characteristics while capturing the quick variance due to the high mobility of vehicular networks. We also propose a reference model to show how the data-driven model can efficiently catch the channel variance and improve the “drive-thru Internet” throughput. Finally, we discuss the opportunities and challenges, which should provide significant reference for the development and standardization of the intelligent LA solution in 802.11 V2X networks.