A Queue Scheduling Method Based on Multiple Priorities

Abstract

The invention discloses a multi-priority-based queue scheduling method, which includes the following steps: classify and process received data packets according to service application types, and classify them into corresponding priority queues; Each priority queue is assigned a weight value; according to the weight value assigned to each priority queue, the corresponding bandwidth value is assigned to each priority queue; each priority queue is polled in order of priority from high to low , and according to the bandwidth allocated by each queue, a corresponding number of data packets are taken out from the corresponding queues for scheduling output. The invention can not only guarantee the service requirement of high-priority business, but also has certain fairness, and can effectively improve the efficiency of queue scheduling. When the service cannot meet the demand, the bandwidth guarantee of real-time or high-priority business can be guaranteed by delaying the demand of low-priority application or reducing the bandwidth of low-priority application. The invention can be widely used in the communication field.

Description

technical field [0001] The invention relates to the technical field of communication, in particular to a multi-priority-based queue scheduling method. Background technique [0002] The existing queue scheduling methods mainly include SP, FIFO, RR, WFQ, etc. The FIFO algorithm schedules according to the arrival time of data packets, that is, first come first served, but this method does not consider the actual business application needs. The strict priority scheduling method of the SP algorithm schedules according to the priority of the queues. After the high priority queues are scheduled, the low priority queues are scheduled, which may cause the low priority queues to be out of service for a long time. The RR algorithm scheduling method schedules multiple queues in proportion. This method has better fairness, but does not consider the application requirements of the business. The WFQ algorithm allocates bandwidth for applications according to the corresponding weight ratio...

AI Technical Summary

Problems solved by technology

[0002] The existing queue scheduling methods mainly include SP, FIFO, RR, WFQ, etc. The FIFO algorithm schedules according to the arrival time of data packets, that is, first come first served, but this method does not consider the actual business application needs

The strict priority scheduling method of the SP algorithm schedules according to the priority of the queue. After the high priority queue is scheduled, the low priority queue is scheduled, which may cause the low priority queue to be out of service for a long time.

The RR algorithm scheduling method schedules multiple queues in proportion. This method has better fairness, but it does not consider the application requirements of the business.

The WFQ algorithm allocates bandwidth for applications according to the corresponding weight ratio. It can effectively guarantee channel bandwidth utilization and zero delay, but cannot guarantee the minimum bandwidth requirements of some multimedia application services.

Images