Research on Edge Resource Allocation Method Based on Vehicle Trajectories Prediction
Shenyang Cao, Yongli Zhang, Ding Pan, Xiaocheng Zhai, Wenwei Liu, Xinyuan Chen, Shuxu Zhao · 2021 IEEE 4th International Conference on Automation, Electronics and Electrical Engineering (AUTEEE) · 2021
Edge computing empowers the Internet of Vehicles to achieve performance requirements such as low latency and high computational complexity for in-vehicle services. Therefore, it is of certain research significance to deploy computing resources along the predicted trajectory of vehicle to improve service quality. Therefore, it is of certain research significance to deploy computing resources along the predicted trajectory of vehicles to improve service quality. However, the randomness and high mobility in the driving of vehicles, and uneven distribution, make the prediction accuracy of vehicle trajectory low, and causes the problems such as unbalanced load of edge servers and low utilization of edge resources. Therefore, the edge resource allocation algorithm based on the improved vehicle trajectory prediction and traffic statistics method is used to obtain the load prediction value of the edge server, then the optimal number of edge resources is calculated on the premise of reducing resource idleness as much as possible. Finally, based on the historical vehicle trajectory data provided by the Didi platform in Chengdu, the use of the edge resource allocation algorithm based on the vehicle trajectory increases the success rate of the request of edge resource from vehicles to about 95%.