Service differentiation for delay-sensitive applications:: an optimisation-based approach

被引:2
|
作者
Key, P
Massoulié, L
Shapiro, JK [1 ]
机构
[1] Univ Massachusetts, Dept Comp Sci, Amherst, MA 01003 USA
[2] Microsoft Res Ltd, Cambridge, England
关键词
congestion control; differentiated service; QoS; pricing;
D O I
10.1016/S0166-5316(02)00118-9
中图分类号
TP3 [计算技术、计算机技术];
学科分类号
0812 ;
摘要
This paper deals with the performance of delay-sensitive applications running over a network that offers multiple classes of service, where the adaption of application rates in response to network feedback is the primary mechanism available for controlling quality of service. We first evaluate the gain in utilisation allowed by the introduction of several classes of service. To this end we compare the pairs of achievable rates, or schedulable regions, for two types of applications with two distinct delay requirements that make use of a single resource, with either no differentiation, simple priority-based differentiation, or earliest-deadline-first (EDF) scheduling-based differentiation. The main observations are that the gain achieved by differentiation is essentially affected by traffic burstiness, and that the two differentiation schemes yield very similar performance. We then consider what feedback information should be sent to traffic sources from different classes, casting the problem in the framework of optimisation-based congestion control. We establish a connection between the sample-path shadow price rationale for feedback synthesis and the rare perturbation analysis technique for gradient estimation in discrete event systems theory. Based on this connection, we propose several marking schemes, for simple priority-based differentiation with a measure of cost based on loss or delay, and also for EDF-based differentiation with loss-based cost. The interaction of these marking algorithms with simple congestion control algorithms is studied via simulations. (C) 2002 Elsevier Science B.V. All rights reserved.
引用
收藏
页码:471 / 489
页数:19
相关论文
共 50 条
  • [1] An efficient quality of service routing algorithm for delay-sensitive applications
    Liu, W
    Lou, WJ
    Fang, YG
    [J]. COMPUTER NETWORKS, 2005, 47 (01) : 87 - 104
  • [2] Supporting Delay-Sensitive IoT Applications: A Machine Learning Approach
    Alnoman, Ali
    [J]. 2020 IEEE CANADIAN CONFERENCE ON ELECTRICAL AND COMPUTER ENGINEERING (CCECE), 2020,
  • [3] How to Network Delay-Sensitive Applications
    Chuprikov, Pavel
    Kogan, Kirill
    [J]. 2021 IFIP NETWORKING CONFERENCE AND WORKSHOPS (IFIP NETWORKING), 2021,
  • [4] A Dynamic Priority Service Provision Scheme for Delay-Sensitive Applications in Fog Computing
    Alnoman, Ali
    Anpalagan, Alagan
    [J]. 2018 29TH BIENNIAL SYMPOSIUM ON COMMUNICATIONS (BSC), 2018,
  • [5] Maximizing User Service Satisfaction for Delay-Sensitive IoT Applications in Edge Computing
    Li, Jing
    Liang, Weifa
    Xu, Wenzheng
    Xu, Zichuan
    Jia, Xiaohua
    Zhou, Wanlei
    Zhao, Jin
    [J]. IEEE TRANSACTIONS ON PARALLEL AND DISTRIBUTED SYSTEMS, 2022, 33 (05) : 1199 - 1212
  • [6] Global and local optimisation-based hybrid approach for cloud service composition
    Shetty, Jyothi
    D'Mello, Demian Antony
    [J]. INTERNATIONAL JOURNAL OF COMPUTATIONAL SCIENCE AND ENGINEERING, 2018, 17 (01) : 1 - 14
  • [7] Distributed multicast routing for delay-sensitive applications
    Im, YJ
    Choi, YH
    [J]. ELECTRONICS LETTERS, 1998, 34 (18) : 1721 - 1722
  • [8] An Optimisation-based Approach for Biopharmaceutical Manufacturing
    Liu, Songsong
    Simaria, Ana S.
    Farid, Suzanne S.
    Papageorgiou, Lazaros G.
    [J]. 24TH EUROPEAN SYMPOSIUM ON COMPUTER AIDED PROCESS ENGINEERING, PTS A AND B, 2014, 33 : 1183 - 1188
  • [9] Controlled loops for quality-of-service differentiation in delay-sensitive optical burst switching networks
    Perello, J.
    Spadaro, S.
    Comellas, J.
    Junyent, G.
    [J]. IET COMMUNICATIONS, 2009, 3 (08) : 1263 - 1273
  • [10] Reinforcement Learning-based Admission Control in Delay-sensitive Service Systems
    Raeis, Majid
    Tizghadam, Ali
    Leon-Garcia, Alberto
    [J]. 2020 IEEE GLOBAL COMMUNICATIONS CONFERENCE (GLOBECOM), 2020,