Research On Resource Management Strategy Based On Vehicular Network With Millimeter Wave

With the development of the fifth-generation communication technology,automatic driving,edge computing,augmented reality and other technologies have come into people’s sight.New broadband services and high-speed massive data pose a huge challenge to the Internet of Vehicles.The maximum transmission rate that traditional Internet of Vehicles technology can provide is limited.Because of the abundant bandwidth and energy concentration,millimeter wave can be used as the key technology to realize Gbps transmission rate.However,electromagnetic waves in the millimeter wave frequency band experience significant attenuation and loss due to their high frequency.Beamforming technology are often used to concentrate the energy of the millimeter wave in a specific direction to compensate for its path loss,ensuring the communication quality of the millimeter wave link.Therefore,millimeter waves have directional characteristics.However,the directional characteristics also bring many challenges to the resource management of millimeter wave communication in vehicular network.This thesis first discusses the characteristics and application scenarios of millimeter wave,and then studies the resource management strategy of the millimeter wave communication in vehicular network for both V2 l and V2 V scenarios.First,in the V2 I scenario,we considers the differential downlink rate requirements of vehicles in the coverage area of the base station and analyzes the impact of millimeter wave beam width on beam alignment time,downlink transmission rate and usage of antenna elements.A beam resource management algorithm based on reinforcement learning is designed to optimize the system downlink throughput on the premise of meeting the limitations on vehicle service requirements,total power and number of antenna components of base stations.The algorithm is divided into two stages: beam selection and joint beam width power adjustment.In the beam selection stage,vehicles are grouped and the beam width adjustment range is limited.Each group is serviced by a beam and a single beam can cover multiple vehicles.In the joint beam width power adjustment stage,under the premise of considering vehicle rate requirements,millimeter wave beam width restrictions,antenna resources restrictions and power resource restrictions,the suboptimal solution of the optimization problem is found by adjusting each beam width and power configuration.The simulation results show that the algorithm can find the beam resource management scheme in many situation.Compared with the fixed beam selection algorithm,the algorithm proposed in this thesis effectively improves the system’s downlink throughput.Secondly,in the V2 V scenario,as a supplement to the V2 I scenario,vehicles have cache capacity.According to the amount of cached resources,vehicles are divided into transmitting vehicles and receiving vehicles.Modeling and analyzing the impact of vehicle location,speed and other factors on the millimeter wave channel capacity,a vehicle association algorithm based on one to many matching is designed for the situation of vehicles cache different service and vehicles demand different services.In this algorithm,the receiving vehicle first determines the optimal beam width of the millimeter wave link,and then the receiving and transmitting vehicle establishes a preference list based on the utility function to match to each other.This thesis simulates and analyzes the influence of some system parameters on the system performance,such as the proportion of transmitting vehicles,the number of quotas and the cycle length.Compared with the fix beam width matching algorithm with 5 or 10 degrees,the cache maximum coincidence matching algorithm and the speed closest matching algorithm,the simulation results verify the advantages of the proposed algorithm in optimizing the transmission capacity of the system.