• DocumentCode
    3550896
  • Title

    Average rate in a M/M/1 processor-sharing queue

  • Author

    Chen, Na ; Jordan, Scott

  • Author_Institution
    Dept. of Electr. Eng. & Comput. Sci., California Univ., Irvine, CA, USA
  • fYear
    2005
  • fDate
    8-10 June 2005
  • Firstpage
    2359
  • Abstract
    We investigate the average rate per job in an open M/M/1 processor-sharing queue. We introduce three definitions of average rate as seen by the system and by each job, and present expressions for each in terms of the system rate and load. We compare the three measures and prove that they are strictly ordered over all loads. We next consider the system rate required to achieve a minimum average rate per job, and prove that it is increasing and convex. Finally, we consider the system rate required to achieve a minimum tail probability on the average rate per job, and present expressions illustrating when the system rate is constrained by the required tail probability.
  • Keywords
    data communication; processor scheduling; queueing theory; M/M/1 processor-sharing queue; minimum average rate per job; minimum tail probability; Context-aware services; Delay; Laplace equations; Operations research; Perturbation methods; Refining; Steady-state; Tail; Time measurement; Traffic control;
  • fLanguage
    English
  • Publisher
    ieee
  • Conference_Titel
    American Control Conference, 2005. Proceedings of the 2005
  • ISSN
    0743-1619
  • Print_ISBN
    0-7803-9098-9
  • Electronic_ISBN
    0743-1619
  • Type

    conf

  • DOI
    10.1109/ACC.2005.1470319
  • Filename
    1470319