• DocumentCode
    3781366
  • Title

    A deterministic optimal task migration algorithm design in NoC-based multi-core system

  • Author

    Fangfa Fu;Jun Liao;Tao Li;Jinxiang Wang

  • Author_Institution
    Microelectronics Center, Harbin Institute of Technology, Harbin, 150001, China
  • fYear
    2015
  • Firstpage
    1
  • Lastpage
    4
  • Abstract
    Task migration is an very important method for the load balancing and fault tolerance in the NoC-based multi-core systems. An global optimal task migration algorithm was proposed to reduce the communication overhead in the process. The algorithm first define the Workload Criterion and give an migration theorem to avoid the migration thrashing and less runtime cost. And then, we adopt the hop area and weigh martrx to find out the optimal paired combination with the minimal migration communication cost. Experimental results show that the proposed algorithm based on the proposed workload criterion reduces migration communication cost approximately by 18% compared with Dynamic Program (DP) algorithm. Moreover, it can get lower energy consumption with an average reduction of 3.9% in our experiments in four networks.
  • Keywords
    "Heuristic algorithms","Approximation algorithms","Multicore processing","Algorithm design and analysis","Fault tolerance","Fault tolerant systems","Runtime"
  • Publisher
    ieee
  • Conference_Titel
    ASIC (ASICON), 2015 IEEE 11th International Conference on
  • Print_ISBN
    978-1-4799-8483-1
  • Electronic_ISBN
    2162-755X
  • Type

    conf

  • DOI
    10.1109/ASICON.2015.7517175
  • Filename
    7517175