Research On Beam Alignment And Content Distribution Scheduling For Millimeter Wave Vehicular Networks

Millimeter wave(mm Wave)communications with abundant spectrum resources can meet the requirements of Intelligent Transportation Systems(ITS)for high throughput and ultralow latency,so that vehicletoeverything communication can be realized.Due to the small antenna aperture at mm Wave frequencies,large antenna arrays are required to provide sufficient link quality.However,searching for the optimal beamforming vectors in large antenna array mm Wave systems requires considerable training overhead,which reduces the system throughput.This is a huge challenge to achieve beam alignment between transceivers and transmitters,especially in the highly dynamic mm Wave vehicular networks environments.At the same time,content distribution plays a crucial role in mm Wave vehicular networks to realize the ITS applications.However,it brings huge challenges to multiuser content distribution in mm Wave vehicular networks due to the high mobility of vehicles and the limited communication resources of roadside units(RSUs).To tackle above challenges,this dissertation focuses on the beam alignment and content distribution scheduling in mm Wave vehicular networks.Firstly,a beam alignment algorithm based on vehicle position information is proposed to achieve fast beam alignment in mm Wave vehicular networks.In the algorithm,RSU obtains a set of candidate beams by obtaining the vehicle position information and training with the double deep Q network(DDQN)algorithm of deep reinforcement learning.Then,according to the criterion of maximizing the system spectral efficiency,the optimal beam of the candidate beam group is obtained by the exhaustive search,so as to achieve fast beam alignment.Performance evaluations demonstrate that the received signaltonoise ratio(SNR)of the vehicle at different positions is greater than the SNR threshold,which avoids the communication interruption and improves the reliability of vehicletoinfrastructure communications.Besides,compared with other search schemes,the proposed scheme attains higher transmission rates,and it is suitable for highspeed mobility mm Wave vehicular networks systems.Secondly,a content distribution scheme based on joint vehicletoinfrastructure(V2I)and vehicletovehicle(V2V)scheduling is proposed for the content distribution problem in mm Wave vehicular networks.When a large number of vehicles send the same content download request to the RSU,the problem of minimizing the total number of content distribution time slots is established from a global optimization perspective while ensuring that all vehicles can complete content downloading.In the V2 I phase,the RSU serially transmits the integrity content to the vehicles,which are selected according to the vehicular network topology.In the V2 V phase,fullduplex communications and concurrent transmissions are exploited to achieve content sharing between vehicles and improve transmission rates.Performance evaluations verify that under the same mm Wave vehicular network topology,the different resource allocation of the RSU will affect the multisource forwarding of content in the V2 V phase,and the content forwarding in the V2 V phase largely determines the effectiveness of resource allocation of the RSU.Performance evaluations demonstrate that the proposed scheme significantly improves the system throughput,especially under largesize file transfers and a large number of vehicles requesting content downloading.