• DocumentCode
    2267407
  • Title

    Study on Parallel Machine Scheduling Problem with Buffer

  • Author

    Li, Shisheng ; Zhou, Yinghua ; Sun, Guangzhong ; Chen, Guoliang

  • Author_Institution
    Univ. of Sci. & Technol. of China, Hefei
  • fYear
    2007
  • fDate
    13-15 Aug. 2007
  • Firstpage
    278
  • Lastpage
    273
  • Abstract
    In the papers by Kellerer et al. (1995) and Zhang (1997) the authors investigate a semi on-line version of a classical parallel machine scheduling problem. In the semi on-line version, there is a buffer of length k which is available to maintain k jobs. The jobs arrive one by one and can be temporarily assigned to the buffer if the buffer is not full. The goal is to assign all jobs to m identical machines such that the makespan is minimized. For two machines, Kellerer et al. (1995) presented a lower bound 4/3 of the worst-case ratio and Zhang (1997) presented an optimal algorithm with a buffer of length 1. In this paper, we investigate the problem with m ges 2 machines. First, we design a new algorithm for Pm//Cmax with a buffer of length lfloorm/2rfloor and analyze it. Second, we present some lower bounds of the worst-case ratio.
  • Keywords
    computational complexity; parallel machines; scheduling; buffers; optimal algorithm; parallel machine scheduling; worst-case ratio; Algorithm design and analysis; Application software; Computer science; Concurrent computing; High performance computing; Parallel machines; Processor scheduling; Scheduling algorithm; Sun;
  • fLanguage
    English
  • Publisher
    ieee
  • Conference_Titel
    Computer and Computational Sciences, 2007. IMSCCS 2007. Second International Multi-Symposiums on
  • Conference_Location
    Iowa City, IA
  • Print_ISBN
    978-0-7695-3039-0
  • Type

    conf

  • DOI
    10.1109/IMSCCS.2007.32
  • Filename
    4392613