Integrated multi-class routing

被引:0
|
作者
Fulgham, ML [1 ]
Snyder, L [1 ]
机构
[1] Univ Washington, Dept Comp Sci & Engn, Seattle, WA 98195 USA
关键词
D O I
暂无
中图分类号
TP301 [理论、方法];
学科分类号
081202 ;
摘要
In this paper we describe a class of routing algorithms called multi-class algorithms. Multi-class algorithms support multiple classes of routing simultaneously, thereby allowing different applications, and even different messages, to select the most advantageous kind of routing. For example some applications prefer the smaller latency variance of oblivious routing, while others prefer the higher throughputs achieved by adaptive routing. Typical systems provide a single class routing algorithm, but applications benefit from the flexibility of multiple classes. Integrated multi-class routers have two characteristics. First, they provide an integrated algorithm where routing classes share resources such as buffers. Each class is not an independent routing algorithm on an independent network, but rather to reduce costs, each class is implemented by a single algorithm on a shared network. Second, multi-class routers help increase performance by providing routing flexibility and network services which help simplify the network interface or system software. The idea of multi-class routing is perhaps obvious and it has appeared before. Our contribution, however, lies in defining multi-class routers, describing their advantages, providing an appropriate method for evaluating such routers, and by demonstrating their usefulness though examples.
引用
收藏
页码:21 / 32
页数:12
相关论文
共 50 条
  • [41] Calibrating Multi-Class Models
    Johansson, Ulf
    Lofstrom, Tuwe
    Bostrom, Henrik
    CONFORMAL AND PROBABILISTIC PREDICTION AND APPLICATIONS, VOL 152, 2021, 152 : 111 - 130
  • [42] On reoptimizing multi-class classifiers
    Chris Bourke
    Kun Deng
    Stephen D. Scott
    Robert E. Schapire
    N. V. Vinodchandran
    Machine Learning, 2008, 71 : 219 - 242
  • [43] Online Multi-Class LPBoost
    Saffari, Amir
    Godec, Martin
    Pock, Thomas
    Leistner, Christian
    Bischof, Horst
    2010 IEEE CONFERENCE ON COMPUTER VISION AND PATTERN RECOGNITION (CVPR), 2010, : 3570 - 3577
  • [44] Learning multi-class dynamics
    Blake, A
    North, B
    Isard, M
    ADVANCES IN NEURAL INFORMATION PROCESSING SYSTEMS 11, 1999, 11 : 389 - 395
  • [45] A review on multi-class TWSVM
    Ding, Shifei
    Zhao, Xingyu
    Zhang, Jian
    Zhang, Xiekai
    Xue, Yu
    ARTIFICIAL INTELLIGENCE REVIEW, 2019, 52 (02) : 775 - 801
  • [46] Calibrated Explanations for Multi-Class
    Lofstrom, Tuwe
    Lofstrom, Helena
    Johansson, Ulf
    13TH SYMPOSIUM ON CONFORMAL AND PROBABILISTIC PREDICTION WITH APPLICATIONS, 2024, 230 : 175 - 194
  • [47] Multi-Class Deep Boosting
    Kuznetsov, Vitaly
    Mohri, Mehryar
    Syed, Umar
    ADVANCES IN NEURAL INFORMATION PROCESSING SYSTEMS 27 (NIPS 2014), 2014, 27
  • [48] Multi-Class Inverted Stippling
    Schulz, Christoph
    Kwan, Kin Chung
    Becher, Michael
    Baumgartner, Daniel
    Reina, Guido
    Deussen, Oliver
    Weiskopf, Daniel
    ACM TRANSACTIONS ON GRAPHICS, 2021, 40 (06):
  • [49] MULTI-CLASS MODEL M
    Emami, Ahmad
    Chen, Stanley F.
    2011 IEEE INTERNATIONAL CONFERENCE ON ACOUSTICS, SPEECH, AND SIGNAL PROCESSING, 2011, : 5516 - 5519
  • [50] Adaptive multi-class controller
    Lo, WL
    Rad, AB
    Tsang, KM
    Wong, YK
    PROCEEDINGS OF THE 2001 IEEE INTERNATIONAL CONFERENCE ON CONTROL APPLICATIONS (CCA'01), 2001, : 18 - 23