A task scheduling method and device for serverless edge computing

Abstract

The embodiment of the present invention provides a task scheduling method and device for serverless edge computing, wherein the method includes: the orchestration control server receives the information of multiple computing functions forwarded by the edge computing server, and based on the priority relationship between the computing functions , construct a directed acyclic graph; for each calculation function in the directed acyclic graph, calculate the delay and energy consumption required to match the calculation function to each candidate server, and obtain the delay and energy corresponding to the task to be processed Among them, the candidate servers include: each edge computing server and the central server in the serverless edge network where the orchestration control server is located; based on the delay and energy consumption corresponding to the tasks to be processed, the target computing server is determined for each computing function. The present invention realizes task scheduling oriented to serverless edge computing.

Description

technical field [0001] The present invention relates to the technical field of big data computing, in particular to a task scheduling method and device for serverless edge computing. Background technique [0002] With the rapid development of new applications such as the Internet of Things, virtual reality, augmented reality, autonomous driving, and interactive games, mobile data generated from various distributed sources shows an explosive growth trend. Cloud computing is used to process such a large amount of data, and cloud computing can better manage and utilize computing resources, network resources, and storage resources. However, since cloud servers are usually deployed in regions far from users, this results in high transmission costs and waiting delays. Edge computing is a decentralized computing architecture. By sinking data caching capabilities, traffic forwarding capabilities, and application service capabilities to edge nodes closer to UE (User End, user termin...

AI Technical Summary

Problems solved by technology

[0003] However, to implement task scheduling in traditional edge computing scenarios, third-party operators need to undertake resource management, scalability processing, load balancing and other responsibilities, and, due to the resource-constrained characteristics of distributed edge infrastructure, edge computing devices The computing power of the remote cloud server is much lower than that of the remote cloud server, which makes the edge computing network have limitations in hosting various applications and services

Images