A multi-resource cloud job scheduling method based on Deep Q-network algorithm

Abstract

The invention provides a multi-resource cloud job scheduling method based on Deep Q-network algorithm, comprising the following steps: collecting current configuration information of resources and demand information of jobs through a cloud environment; respectively representing The current configuration information of the resource and the demand information of the job by a matrix image, the cellsincluding the cells, the cells of the same color represent the same job, and the rectangles formed by the cells of the same color comprise M *N cells, M represents the number of resources and N represents the time step; the rectangles formed by the cells of the same color comprise M* N cells. According to the matrix image, obtaining high-level semantic information by using a depth learning method;According to the high-level semantic information, a reinforcement learning method is used to complete the real-time scheduling planning of resources. A matrix image of thepresent invention can completely and clearly represent the state of resource and jobs. The method also uses depth reinforcement learning, which combines depth learning with reinforcement learning, in which depth learning mainlycompletes state perception, while reinforcement learning completes decision-making and realizes state-to-action mapping.

Description

technical field [0001] The invention relates to the field of cloud computing resource scheduling, in particular to a multi-resource cloud job scheduling method based on Deep Q-network algorithm. Background technique [0002] Resource scheduling is a research difficulty and hotspot in the field of cloud computing. A good resource allocation and scheduling strategy can effectively utilize resources and increase the economic benefits of suppliers while ensuring the quality of service (QoS) for users. Cloud computing resource scheduling is actually a multi-constraint, multi-objective optimization NP-hard problem. The current traditional method to solve the decision-making problem is to design an efficient heuristic algorithm with performance guarantee under specific conditions, which is not universal and practical, and cannot adapt to the changing and complex cloud environment. In addition, the researchers abstracted resource scheduling in the cloud environment into a sequenti...

AI Technical Summary

Problems solved by technology

The disadvantage of this patent is that the active virtual machine data cannot fully represent the status of resources and jobs in the cloud environment

Images