A Double Layer Isolation Method of Network Resources Based on Programmable Switch

Abstract

The invention discloses a double-layer isolation method for network resources based on a net field programmable gate array. In the method, the network resources are divided into a gathered flow layer and a flow layer, a specified bandwidth is reserved for each gathered flow when the gathered flow layer is scheduled, and a plurality of flows in each gathered flow compete for the bandwidth according to the weight when the flow layer is scheduled. By the aid of a low-cost Net FPGA (field programmable gate array) hardware device, the network resources are flexibly, fairly and controllably distributed, network resource distribution implementation cost is reduced, the implementation method is simple and convenient, measures are flexible, and network service quality can be remarkably ensured and is unrelated to a network layer protocol.

Description

technical field [0001] The invention relates to the technical field of computer networks, in particular to a method for double-layer isolation of network resources based on a programmable switch (Net Field Programmable Gate Array, NetFPGA). Background technique [0002] The Internet has achieved great success in the past few decades, and it provides people with a brand-new platform for exchanging information and receiving information. Its widespread popularity has promoted the development and application of network technology. However, the huge network also limits its future development. For example, with the scale and size of the current network, it is very difficult to guarantee the quality of service for each network service. Traditional QoS implementation methods are generally based on integrated service or differentiated service models. Integrated services reserve separate resources for each business, but it is easy to waste resources; differentiated services flexibl...

AI Technical Summary

Problems solved by technology

Integrated services reserve separate resources for each business, but it is easy to cause waste of resources; differentiated services flexibly allocate resources according to labels such as priority, but cannot reserve isolated network resources for a single business

[0003] The models of integrated services and differentiated services have their own advantages, but they lack flexibility and cannot adapt to changing network requirements; on the other hand, the current network resource allocation methods are mostly based on software simulation or theoretical derivation, and lack specific implementation methods

Images