Software defined network (SDN) congestion control routing method based on minimization of path overhead and stream scheduling cost

Abstract

The invention relates to a software defined network (SDN) congestion control routing algorithm based on minimization of a path overhead and a stream scheduling cost. A controller performs periodic monitoring on use ratios of all links connected with an interchanger controlled by the controller, wherein the formula of the use ratio is shown as in the description, when the use ratio is greater than or equal to a certain fixed value Etath, it is determined that the link is congested, and the congestion control routing algorithm is used. The algorithm comprises steps of determining all large streams on the links according to the stream rate of the congested link so as to reserve the large streams for scheduling; performing rerouting computation on all large streams on the links, computing the overhead of each path, and selecting a path of minimum overhead from a plurality of equivalent paths to be used as a scheduling path; and finally, computing the scheduling cost of each stream, and selecting the stream of the minimum cost to schedule. The algorithm not only can schedule the congested link effectively, but also makes the network link resource more efficient, improves the use ratio of the link and shortens the stream transmission time.

Description

technical field [0001] The invention relates to the field of routing algorithms of SDN networks. In particular, this algorithm is based on the network-wide network status information view of the SDN network controller, and is mainly designed and optimized for application scenarios (such as data center networks) in which the amount of data in the SDN network is large, the traffic is highly dynamic, and it is easy to cause network congestion. The purpose of the routing algorithm is to quickly relieve network congestion while increasing network resource utilization. Background technique [0002] In recent years, with the development of social informatization, people's demand for the Internet has shown explosive growth. In addition to the scale and development speed of the Internet far exceeding all previous data networks, the speed of business innovation is also unprecedentedly accelerated. Various real-time services such as video Voice, cloud data centers, and mobile services...

AI Technical Summary

Problems solved by technology

According to the basic characteristics of the flow, the flow is divided into large flow and small flow. In the scheme proposed by KanagaveluR et al. in the article "OpenFlowbased control for re-routing with differentiated flows in DataCenterNetworks", in the process of selecting the scheduling flow, the data center network traffic has the characteristics of large and small flows. Size classification, and only scheduling large flows, improves the effectiveness of rerouting, but at the same time it does not conduct further research on the selection of scheduled large flows, only selects the largest flow for scheduling, taking into account the characteristics of the flow and the entire network Effective use of resources, the maximum flow is not necessarily the best scheduling flow, which makes the selected flow not optimal

Images