首页> 外文会议>High-Performance Computing, 1997. Proceedings. Fourth International Conference on >Probabilistic routing in wavelength-routed multistage, hypercube,and Debruijn networks
【24h】

Probabilistic routing in wavelength-routed multistage, hypercube,and Debruijn networks

机译:波长路由多级超立方体中的概率路由和Debruijn网络

获取原文

摘要

Optical networks based on wavelength division multiplexing (WDM)and wavelength routing are considered to be potential candidates for thenext generation of wide area networks. One of the main issues in thesenetworks is the development of efficient routing algorithms whichrequire a minimum number of wavelengths. We focus on the permutationrouting problem in multistage WDM networks which we call 2-multinets. Wepresent a simple, oblivious probabilistic approach which solves thepermutation routing problem on 2-multinets with very high probability(in the usual theoretical sense) using O(log2N/loglogN)wavelengths, where N is the number of nodes in the network, therebyimproving the previous result due to Pankaj and Gallager (1995) thatrequires O(log3 N) wavelengths. Our approach is advantageousand practical as it is simple, oblivious, and suitable for centralizedas well as distributed implementations. We also note that O(logN)wavelengths will suffice with good probabilistic guarantee for the caseof dynamic permutation routing where requests arrive and terminatewithout any relation to each other. The above results are for networkswith wavelength converters and we show that the use of converters can beeliminated at the expense of a factor of log N more wavelengths. We alsoshow how our approach can be used to solve the dynamic permutationrouting problem well (in practice), using O(1) wavelengths on thehypercube and O(logN) wavelengths on the Debruijn network. These improvethe previous known bounds of O(logN) and O(log2N),respectively
机译:基于波分复用(WDM)的光网络 和波长路由被认为是潜在的候选者 下一代广域网。这些主要问题之一 网络是高效路由算法的发展, 需要最少数量的波长。我们专注于排列 多级WDM网络中的路由问题,我们称为2-multinet。我们 提出了一种简单的,遗忘的概率方法,该方法解决了 2多网上的高置换路由问题 (在通常的理论意义上)使用O(log 2 N / loglogN) 波长,其中N是网络中的节点数,因此 由于Pankaj和Gallager(1995)改善了先前的结果, 需要O(log 3 N)个波长。我们的方法是有优势的 简单实用,易于操作,适用于集中式 以及分布式实现。我们还注意到O(logN) 波长将满足情况的良好概率保证 请求到达和终止的动态排列路由的说明 彼此之间没有任何关系。以上结果适用于网络 波长转换器,我们证明了使用转换器可以 消除的代价是增加了N个对数波长。我们也 展示如何使用我们的方法来解决动态排列 很好地解决路由问题(在实践中),在O上使用O(1)波长 Debruijn网络上的超立方和O(logN)波长。这些改善 O(logN)和O(log 2 N)的先前已知范围, 分别

著录项

相似文献

  • 外文文献
  • 中文文献
  • 专利
获取原文

客服邮箱:kefu@zhangqiaokeyan.com

京公网安备:11010802029741号 ICP备案号:京ICP备15016152号-6 六维联合信息科技 (北京) 有限公司©版权所有
  • 客服微信

  • 服务号