输入队列交换的公平可扩展调度系统

Fair scalable scheduling architectures for input-queued switches

提出了一种基于输入队列交换的公平可扩展网络调度系统FSSA。通过将若干个容量较小的调度器合理连接并使其协同工作,构成多端口大容量网络交换调度系统,解决了单个调度器容量和端口数受集成电路工艺限制的问题。FSSA不仅速度高、规模可扩展而且易于硬件实现。环型连接、管线工作及公平调度技术的采用使FSSA在性能方面得到了进一步优化。仿真结果显示,FSSA的性能可与基于iSLIP、DSRR等算法的单片调度器相比拟,尤其在流量较大时,FSSA的性能明显优于单调度器性能。

This paper presents a fair scalable scheduling architecture （FSSA） for input queued switches. In the FSSA, a large scheduler is realized by connecting several small scheduling devices and making them work in fair pipeline scheduling. The important feature of this architecture is that it can be configured as a large scheduler with high performance but without the capacity limitations due to the size of monolithic scheduling device, The simulation results show that the performance of the FSSA is better than or very close to that of other typical scheduling algorithms under high traffic load.