具有对数延迟的输入队列交换机:必要条件和可重构调度算法

Krishnendu Roy, R. Vaidyanathan, J. Trahan
{"title":"具有对数延迟的输入队列交换机:必要条件和可重构调度算法","authors":"Krishnendu Roy, R. Vaidyanathan, J. Trahan","doi":"10.1145/1477942.1477961","DOIUrl":null,"url":null,"abstract":"Typically, a scheduling algorithm for an <i>n</i> x <i>n</i> packet switch with a crossbar as the data fabric divides time into slots, each of duration <i>t<sub>p</sub></i> sufficient to transmit a packet. If a scheduling round requires <i>t<sub>r</sub></i> > <i>t<sub>p</sub></i> time, then the switch can transmit multiple packets, up to <i>s</i> = ⌊<i>t<sub>r</sub>/t<sub>p</sub></i>⌋, between each mapped input-output pair under the current mapping. If <i>s</i> = 1, there exists a frame-based scheduling algorithm with Θ(log <i>n</i>) delay. For uniform random traffic, we establish that the delay is Ω(<i>n</i>) for any <i>s</i> > 1, hence, <i>s</i> = 1 is the only case where a Θ(log <i>n</i>) delay is achievable.\n Given the importance of achieving a low <i>s</i>, it is imperative to develop extremely fast scheduling algorithms (that reduce <i>t<sub>r</sub></i>) on a mesh-based structure (corresponding to the crossbar topology of the switch). We present results for a fast scheduling algorithm that runs on a mesh-of-trees topology that can be overlaid on the crossbar switching fabric.","PeriodicalId":329300,"journal":{"name":"Symposium on Architectures for Networking and Communications Systems","volume":"42 1","pages":"0"},"PeriodicalIF":0.0000,"publicationDate":"2008-11-06","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":"{\"title\":\"Input-queued switches with logarithmic delay: necessary conditions and a reconfigurable scheduling algorithm\",\"authors\":\"Krishnendu Roy, R. Vaidyanathan, J. Trahan\",\"doi\":\"10.1145/1477942.1477961\",\"DOIUrl\":null,\"url\":null,\"abstract\":\"Typically, a scheduling algorithm for an <i>n</i> x <i>n</i> packet switch with a crossbar as the data fabric divides time into slots, each of duration <i>t<sub>p</sub></i> sufficient to transmit a packet. If a scheduling round requires <i>t<sub>r</sub></i> > <i>t<sub>p</sub></i> time, then the switch can transmit multiple packets, up to <i>s</i> = ⌊<i>t<sub>r</sub>/t<sub>p</sub></i>⌋, between each mapped input-output pair under the current mapping. If <i>s</i> = 1, there exists a frame-based scheduling algorithm with Θ(log <i>n</i>) delay. For uniform random traffic, we establish that the delay is Ω(<i>n</i>) for any <i>s</i> > 1, hence, <i>s</i> = 1 is the only case where a Θ(log <i>n</i>) delay is achievable.\\n Given the importance of achieving a low <i>s</i>, it is imperative to develop extremely fast scheduling algorithms (that reduce <i>t<sub>r</sub></i>) on a mesh-based structure (corresponding to the crossbar topology of the switch). We present results for a fast scheduling algorithm that runs on a mesh-of-trees topology that can be overlaid on the crossbar switching fabric.\",\"PeriodicalId\":329300,\"journal\":{\"name\":\"Symposium on Architectures for Networking and Communications Systems\",\"volume\":\"42 1\",\"pages\":\"0\"},\"PeriodicalIF\":0.0000,\"publicationDate\":\"2008-11-06\",\"publicationTypes\":\"Journal Article\",\"fieldsOfStudy\":null,\"isOpenAccess\":false,\"openAccessPdf\":\"\",\"citationCount\":\"0\",\"resultStr\":null,\"platform\":\"Semanticscholar\",\"paperid\":null,\"PeriodicalName\":\"Symposium on Architectures for Networking and Communications Systems\",\"FirstCategoryId\":\"1085\",\"ListUrlMain\":\"https://doi.org/10.1145/1477942.1477961\",\"RegionNum\":0,\"RegionCategory\":null,\"ArticlePicture\":[],\"TitleCN\":null,\"AbstractTextCN\":null,\"PMCID\":null,\"EPubDate\":\"\",\"PubModel\":\"\",\"JCR\":\"\",\"JCRName\":\"\",\"Score\":null,\"Total\":0}","platform":"Semanticscholar","paperid":null,"PeriodicalName":"Symposium on Architectures for Networking and Communications Systems","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/10.1145/1477942.1477961","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"","JCRName":"","Score":null,"Total":0}
引用次数: 0

摘要

通常情况下,使用交叉条作为数据结构的n × n分组交换机的调度算法将时间划分为插槽,每个插槽的持续时间为tp,足以传输数据包。如果一个调度轮需要tr > tp的时间,则在当前映射下,交换机可以在每个映射的输入输出对之间传输最多s =⌊tr/tp⌋的数据包。当s = 1时,存在时延为Θ(log n)的基于帧的调度算法。对于均匀随机流量,我们建立了对于任何s > 1的延迟为Ω(n),因此,s = 1是唯一可以实现Θ(log n)延迟的情况。考虑到实现低s的重要性,在基于网格的结构(对应于交换机的横杆拓扑)上开发极快的调度算法(减少tr)是势在必行的。我们给出了一种快速调度算法的结果,该算法运行在树状网格拓扑结构上,可以覆盖在交叉交换结构上。
本文章由计算机程序翻译,如有差异,请以英文原文为准。
Input-queued switches with logarithmic delay: necessary conditions and a reconfigurable scheduling algorithm
Typically, a scheduling algorithm for an n x n packet switch with a crossbar as the data fabric divides time into slots, each of duration tp sufficient to transmit a packet. If a scheduling round requires tr > tp time, then the switch can transmit multiple packets, up to s = ⌊tr/tp⌋, between each mapped input-output pair under the current mapping. If s = 1, there exists a frame-based scheduling algorithm with Θ(log n) delay. For uniform random traffic, we establish that the delay is Ω(n) for any s > 1, hence, s = 1 is the only case where a Θ(log n) delay is achievable. Given the importance of achieving a low s, it is imperative to develop extremely fast scheduling algorithms (that reduce tr) on a mesh-based structure (corresponding to the crossbar topology of the switch). We present results for a fast scheduling algorithm that runs on a mesh-of-trees topology that can be overlaid on the crossbar switching fabric.
求助全文
通过发布文献求助,成功后即可免费获取论文全文。 去求助
来源期刊
自引率
0.00%
发文量
0
×
引用
GB/T 7714-2015
复制
MLA
复制
APA
复制
导出至
BibTeX EndNote RefMan NoteFirst NoteExpress
×
提示
您的信息不完整,为了账户安全,请先补充。
现在去补充
×
提示
您因"违规操作"
具体请查看互助需知
我知道了
×
提示
确定
请完成安全验证×
copy
已复制链接
快去分享给好友吧!
我知道了
右上角分享
点击右上角分享
0
联系我们:info@booksci.cn Book学术提供免费学术资源搜索服务,方便国内外学者检索中英文文献。致力于提供最便捷和优质的服务体验。 Copyright © 2023 布克学术 All rights reserved.
京ICP备2023020795号-1
ghs 京公网安备 11010802042870号
Book学术文献互助
Book学术文献互助群
群 号:481959085
Book学术官方微信