Network resource scheduling method and server

Abstract

The invention discloses a network resource scheduling method and a server. Wherein in the network resource scheduling method, for the resource request sent by at least one user received within a specified time slice, the resource request is assigned to a subqueue associated with the service type of the resource request according to the service type of the resource request. Calculate the consumption rate of the business associated with the jth sub-queue on the i-th resource segment, and predict that within a predetermined time slice in the future, the business associated with the j-th sub-queue will be on the i-th resource segment The predicted consumption rate C'ij of . The current status information of network resources is collected, and the resource requests in each sub-queue are scheduled by using the predicted consumption rate and the current status information of network resources. By predicting the degree of congestion of network resources, the rational and fair use of end-to-end network resources is guaranteed, and the overall throughput of the operation network and business system is improved.

Description

technical field [0001] The invention relates to the communication field, in particular to a network resource scheduling method and server. Background technique [0002] Improving user experience is an end-to-end performance optimization process in a network environment, involving bandwidth and QoS (Quality of Service) in the access network, backbone network, and IDC (Internet Data Center, referred to as: Internet Data Center), etc. The guarantee capability of different network resources needs to allocate the resource requirements of users / services hop by hop, reasonably guarantee and prevent abuse. During the access process from the user terminal to the server, the network resources required by the business are distributed, and the required service levels are also differentiated. During the successive allocation / occupancy of resources at each hop, if a large number of concurrent user resource requests cannot be synchronized Scheduling, the utilization rate of network resour...

AI Technical Summary

Problems solved by technology

Within a time slice, the above business requests must be processed in time and resource scheduling is performed as needed. If scheduling is only performed in order of arrival time, some services will compete for local / end-to-end link channels (bandwidth resources, or different QoS level channel) phenomenon, resulting in link congestion or uneven load, ultimately reducing the overall network carrying efficiency

[0004] Ordinary network resource scheduling algorithms are used to schedule a single network resource. When used in a multi-hop resource environment, the complexity is high and the delay is large. The resource scheduling method of come-first-serve has a low utilization rate of network resources when scheduling in distributed multi-hop resources

Images