具有優(yōu)先服務(wù)機(jī)制的嵌套式DRR算法
A Prioritized Nested DRR Algorithm
-
摘要: 針對DRR算法在保證業(yè)務(wù)時(shí)延性能上的固有不足,該文提出了具有優(yōu)先服務(wù)策略的嵌套式DRR算法,該算法對需要保證延時(shí)的業(yè)務(wù)類實(shí)施專門的服務(wù)策略,利用漏桶控制和虛令牌的分配機(jī)制,并改善服務(wù)隊(duì)列的調(diào)度策略,在不增加算法復(fù)雜度的情況下,有效地減小了時(shí)延敏感業(yè)務(wù)隊(duì)列中包的延遲。文章證明了算法在獲得延時(shí)保證的同時(shí),選擇合適的參數(shù),仍然能夠維持調(diào)度算法的公平性。仿真結(jié)果表明算法對時(shí)延性能的改善是明顯的。
-
關(guān)鍵詞:
- 調(diào)度; DRR算法; 延時(shí)特性; 公平性
Abstract: As DRR fails to provide strong latency bound, in this paper, a new scheduling discipline named Prioritized Nested DRR (PNDRR) is presented, which introduces a token bucket with virtual allocated token quantum and changes the entering order of the scheduling list for the latency-critical flow. By using the scheme for the latency critical flow, delay of packet in latency critical queue is effectively diminished. In this paper, theoretical analyses prove that PN DRR results in a significant improvement in the latency bound of delay-sensitive traffic in comparison to nested DRR(NDRR) while preserves the good properties as the same relative fairness bound and the per-packet complexity of 0(1) as NDRR. Simulation results also support analysis. -
Parekh A K, Gallager R G. A generalized processor sharing approach to flow control in integrated services networks: The single node case[J].IEEE/ACM Trans. on Networking.1993, 1(3):344-[2]Shreedhar M, Varghese G. Efficient fair queueing using deficit round robin[J].IEEE/ACM Trans. on Networking.1996, 4(3):375-[3]Macgregor M, Shi W. Deficit for bursty latency-critical flows:DRR++, IEEE International Conference on Networks, Singapore,2000:287 - 293.[4]Kanhere S S, Sethu H. Fair, efficient and low-latency packet scheduling using nested deficit round robin. 2001 IEEE Workshop on High Performance Switching and Routing,, Texas,USA, 2001:6 - 10.Mao J M, Moh W M, Wei B. PQWRR scheduling algorithm in supporting of DiffServ. International Communication Confrence,Amsterdam, the Netherlands, 2001:679 - 684.[5]Stiliadis D.[J].Varma A. Latency-rate servers: A general model for analysis of traffic scheduling algorithms. Proceedings of IEEE INFOCOM96, San Francisco, USA.1996,:- -