Task migration decision-making method, device and system

Abstract

The invention provides a task migration decision-making method, device and system. The method comprises the steps of: acquiring local experience information of at least one time period uploaded by a plurality of terminals, integrating a plurality pieces of local experience information of the same time period, and generating global experience information of the time period; selecting training parameters of different decision groups from the global experience information of a plurality of time periods for integration, generating training experience information, training all online execution networks according to the training experience information, and determining network parameters of each online execution network; and sending the network parameters corresponding to a decision group to terminals in the decision group according to the grouping information of the terminals. According to the task migration decision-making method, device and system provided by the embodiment of the invention, network parameters of a large number of terminals can be updated, the training stability and the training result performance are improved, and the task migration decision-making effect of the terminals can be ensured.

Description

Technical field [0001] The present invention relates to the field of wireless communication, and in particular, to a method, apparatus, and system of task migration decisions. Background technique [0002] With the development of mobile communication technology, especially the arrival of the 5G era, the number of mobile devices is expected to have explosive growth. At the same time, due to the development of terminal equipment and the increase in user needs, some computational intensive tasks should also be born, such as enhance reality (AR), virtual reality (VR), online artificial intelligence and Internet of Things. These new tasks typically need to process a large amount of data and sensitive to delay, while terminal devices are generally restricted low power devices. It is difficult to meet the limited computing power of the terminal local local local local local local, so people propose the method of moving the edge calculation to solve the problem of resource resource resou...

AI Technical Summary

Problems solved by technology

[0006] However, neither the traditional algorithm nor the deep reinforcement learning method considers the cooperation and competition between multiple users when learning the migration strategy. It is not necessarily optimal in terms of the total overhead (delay and power consumption) of the system.

When applied to a NOMA (non-orthogonal multiple access)-based MEC system, due to severe interference between users, its migration algorithm may not be able to effectively make decisions or even cause training instability

Images