A kind of intelligent qos method and system based on token bucket

Abstract

The invention provides an intelligent QOS method and system based on a token bucket, belonging to the technical field of network applications. The method of the present invention comprises the steps of: marking the priority of the message, and then sending the data packet of the marking priority into the user guarantee token bucket; judging whether it is within the guarantee bandwidth, if yes, sending the data packet into the system guarantee queue, if not , send the data packets into the user queue according to the user classification; the scheduler prioritizes the data packets in the system guarantee queue into the system token bucket according to the priority order, and transfers the data packets in the user queue into the system token bucket if there is remaining bandwidth ; The scheduler sends packets according to the tokens in the system token bucket. The beneficial effect of the present invention is: the system adds a penalty mechanism on the principle of maintaining fairness, so that some applications no longer occupy a large amount of bandwidth, and the bandwidth is allocated to other applications, and the delay is significantly reduced, and the user's Internet access environment is significantly improved.

Description

Technical field[0001]The present invention relates to the technical field of network applications, in particular to a smart QOS method based on a token bucket, and also to a system for realizing the smart QOS method.Background technique[0002]With the development of IP technology and networks, in the past, all the messages in the network were treated indiscriminately, first-in-first-out, and unreliable transmission methods such as first-in-first-out and best-effort delivery of messages were no longer adapted to the various applications emerging in the current network. , Video, voice and other complex network services.[0003]Traditional QOS is mostly based on a stream (a data stream identified by a quintuple) for queuing and congestion processing flow control, without distinguishing between different application services, and it cannot be achieved under the condition of limited total network resources. In order to ensure the bandwidth of some applications, the quality of service of oth...

AI Technical Summary

Problems solved by technology

[0002] With the development of IP technology and network, the unclear and unreliable transmission methods such as treating all packets in the network indiscriminately, adopting first-in-first-out, and best-effort delivery in the past are no longer suitable for various applications emerging in the current network. , video, voice and other complex network services

[0003] Traditional QOS is mostly based on a flow (data flow identified by quintuple) to perform flow control of queuing and congestion processing, without distinguishing different application services, so it cannot be realized in the case of limited total network resources. In order to guarantee the bandwidth of some applications and reduce the quality of service requirements of other applications

[0004] In the daily office process of an enterprise, different employees have different requirements for network applications. Usually, irrelevant businesses occupy a large amount of bandwidth resources. When the network is overloaded or congested, the data packets of the core business have no corresponding As a result, the bandwidth guarantee is delayed or discarded. Since the total bandwidth is fixed, to solve the problem of bandwidth resource preemption caused by business needs, it is necessary to propose a set of reasonable allocation of bandwidth resources for each user. Effectively manage Internet traffic, improve network service quality, and dynamically and reasonably allocate bandwidth

Images