Dynamic grouping Internet of Vehicles caching method based on reinforcement learning in MEC environment

Abstract

The invention discloses a dynamic grouping Internet of Vehicles caching method based on reinforcement learning in an MEC environment. The method comprises the following steps: S1, obtaining all to-be-grouped vehicle information, road side unit information, macro base station information, a core network and caching content information; s2, determining a mobile vehicle group cache pool and a central vehicle; s3, constructing a system content cache and delivery model according to the mobile vehicle group cache pool, the road side unit and the macro base station information; s4, establishing a multi-agent collaborative edge caching model based on an actor-commentator framework by taking the minimum total delay of the acquired contents of all the vehicles in the vehicle group as an objective function; and S5, obtaining a corresponding caching strategy through the training model, and selecting a strategy with the minimum system delay to perform content caching. The problems of overlarge load of the base station and overlong delay of end-to-end content acquisition caused by frequent communication between the vehicle and the roadside base station can be solved.

Description

technical field [0001] The present invention relates to mobile edge computing technology and the field of Internet of Vehicles, in particular to a dynamic grouping Internet of Vehicles caching method based on reinforcement learning in an MEC environment, and in particular to a dynamic grouping—grouping vehicles traveling in different directions to form a dynamic cache Pool, combined with multi-agent reinforcement learning to get the optimal caching strategy. Background technique [0002] The Internet of Vehicles has become an important application scenario under 5G communication technology. The execution of vehicle-related applications is supported through the Internet of Vehicles. These applications require a stable communication environment and low-latency data transmission, which proposes large-scale links and high throughput for the construction of the Internet of Vehicles. Relevant needs for volume support. In order to ensure the safety and accuracy of vehicle applicat...

AI Technical Summary

Problems solved by technology

[0005] The purpose of the present invention is to propose a dynamic grouping Internet of Vehicles cache method based on reinforcement learning in the MEC environment in order to solve the problem of excessive delay caused by network transmission pressure, which can alleviate the gap between the edge server and the central cloud in the prior art. Transmission and storage pressure, reduce transmission delay

Images