Edge computing user task unloading method

Abstract

The invention discloses an edge computing user task unloading method in the field of mobile communication. The task unloading problem for a user to a base station in the dense cellular network is notonly influenced by environmental noise, but also influenced by same-channel interference. According to the method, a user task unloading condition when the same-channel interference exists in the heterogeneous network is considered; a communication model is established. An algorithm based on numerical estimation is proposed to estimate the concept of the initial temperature of a simulated annealing algorithm. The approximate initial temperature is searched by adopting variable-step-length numerical estimation. The problem that algorithm time is too long due to the fact that random initial temperature is used is solved, then the simulated annealing algorithm is used for circularly changing a target server of a user, the searching process of a solution is controlled through a cooling function combining logarithms and polynomials, and a task unloading scheme enabling the average unloading time delay of the user to be minimum is searched for. Compared with a traditional method, the methodcan obtain an optimization scheme more quickly, and has better practicability.

Description

technical field [0001] The invention belongs to the technical field of mobile communication, and relates to a method for unloading tasks of edge computing users. Background technique [0002] Network applications require more and more data, and the real-time requirements are also getting higher and higher. The traditional data request method of directly connecting the client to the cloud server becomes difficult to meet the quality of service (Quality of Service, QoS). The main bottleneck lies in the large amount of data. For the traffic pressure caused by the backhaul link of the core network, and the delay problem caused by too many data transmission links. [0003] Mobile devices have more and smarter service functions, which means that more power consumption is required to support higher processing capabilities, and are more indispensable in real life, which makes the frequency and time of mobile devices used are very different. However, since the most important advanta...

AI Technical Summary

Problems solved by technology

[0005] In order to make full use of the computing power of the edge server and alleviate its own energy consumption pressure, mobile devices choose to upload tasks to the MEC server for computing. The fast computing power of the MEC server can shorten the time, but it takes time to upload tasks, and multiple user tasks queue up on the MEC server. Will cause delay, so a good task scheduling method is required

Images