Research On Wireless Resource Allocation Algorithm Based On 5G Vehicle Network

Advanced intelligent driving requires high speed and low delay.Internet of vehicles is known as the most promising communication technology in intelligent driving,and resource allocation plays a crucial role on the Internet of Vehicles communication.This thesis mainly focuses on the research and analysis of vehicle equipment autonomy in the selection of spectrum resources in mode 2.The influence of the proposed resource allocation method on system reliability and delay are verified from the perspective of simulation by evaluating the average packet rate and average packet time interval between the proposed resource allocation scheme and the supported schemes in the existing standards.The work and innovation of the thesis are as follows:1.Due to the influence of vehicle half-duplex working mode,the loss of vehicle sensing scheduling information in mode 2 causes the problem of inaccurate interference measurement.Therefore,a resource allocation method based on the combination of vehicle clustering and perception is proposed.All vehicles in the scene first access the Roadside unit with the closest distance according to their own geographical position.Vehicles in the same driving direction and Roadside unit are divided into a cluster.Then,the exclusive time-domain resources are divided for the vehicle cluster,and the timedomain resources between different vehicle cluster are orthogonal.Finally,cluster members,based on their own perception,upload the results to the header car and coordinate resources by the header car.Simulation results show that clustering and exclusive resource division can reduce inter-vehicle interference in the time domain,and head-car management can reduce the probability of failure of perceptual scheduling task and improve the reliability of the system.2.In aperiodic service,the packet arrival time is not fixed,the vehicle cannot predict the resource interference problem,so two resource allocation mechanisms of autonomous resource selection are proposed: the first is to monitor the Scheduling Assignment in real time.In view of the monitoring results,the reserved resources are deactivated or activated,and the vehicle can take the deactivated reserved resources as candidate resources;The second is based on the resource allocation mechanism of Listen Before Talk.Each vehicle initializes a counter at the time of packet arrival,measures the interference on the current transmission slot to determine whether the counter is reduced.When the counter is less than or equal to zero,the packet is issued on the next slot of the current measured slot.Simulation results show that the above two mechanisms can improve the reliability of the system.3.The vehicle carries out data transmission in the way of resource reservation.When the burst packet arrives,there are no available resources,which results in packet loss,So that a resource competition mechanism based on dual perception window is proposed.First of all,vehicle can blind detection the resource Scheduling Assignment of other vehicles in the large-scale sensing window,and the semi-static resources are reserved according to the blind detection results.When a package burst arrival and the delay of reserved resources exceeds the maximum delay,the vehicle can perform small-scale sensing.Priority determination is made according to the perceived business type,or time slot resource occupation is determined according to the size of data volume in the smallscale sensing window.if lower than its own traffic priorities or reserved multiple resources but use only a part of them,vehicle can use low priority traffic reserved resources or unused parts.Simulation results show that this strategy can improve the average packet collection rate and reduce the average packet collection delay in the aperiodic traffic.