Research On Millimeter-Wave Beam Tracking Strategies For V2I Communication In Vehicular Networks

Vehicles can obtain a large amount of data from edge servers or the Internet through V2I(Vehicle to Infrastructure)which is one of the most basic and important communication methods in V2X(Vehicle to Everything).With the development of 5G communication networks,the system throughput of the network has attracted much attention since more and more data needs to be obtained from base stations or roadside units.Millimeter wave is a candidate band for 5G due to its extremely high spectrum resources and it is usually combined with beamforming technology to improve system throughput.However,due to the high-speed mobility of the vehicle in IoV(Internet of Vehicles)scenario,the antenna beam between the vehicle and the base station(or roadside unit)needs to be frequently aligned,resulting in a large amount of beam training overhead to ensure the quality of the communication link.In order to reduce the beam training overhead,beam switching technology which can sequentially switch the beams using the vehicle’s position information to achieve the purpose of beam tracking can be used.In the beam switching scheme,the beam width will directly affect the channel gain of the antenna,which will have a crucial impact on the communication quality of the IoV and optimizing the beam width design can improve the system performance.This article first introduces the basic principles of IoV and millimeter wave beam tracking.Then,in order to solve the problems of system interruption and system throughput degradation due to inaccurate vehicle position prediction in single-vehicle V2I communication,two beam tracking strategies,namely beam hard switching and beam relay switching based on vehicle position prediction,and their corresponding beam design schemes are proposed.The simulation results show that the average data transmission rate of the two design schemes proposed in this paper is improved by about 50%compared with the existing uniform angle beamforming scheme.When comparing the two proposed schemes,we can find that the beam relay switching tracking scheme can reduce the interruption rate to 50%of the beam hard switching tracking scheme at the expense of energy consumption,and its average data transmission rate is slightly better than the beam hard handover tracking scheme.Secondly,in order to solve the problems of uneven distribution of beam resources between vehicles and system throughput degradation due to changes in vehicle node distribution in multi-vehicle V2I communication,a beam design scheme based on vehicle distribution is proposed.The beam design problem is then modeled as a Markov decision process and solved by using a deep reinforcement learning algorithm.When optimizing the beam width,not only the throughput of the system,but also the fairness between vehicles is considered.Simulation experiments show that the proposed scheme is significantly superior to the existing uniform coverage beamforming and uniform angle beamforming schemes,and the algorithm can well adapt to changes in the distribution of vehicle nodes,it converges fast and achieves near-optimal performance.Finally,this article puts forward a preliminary idea for the future research work.