• DocumentCode
    1934721
  • Title

    Power-Aware Control Flow Checking Compilation: Using Less Branches to Reduce Power Dissipation

  • Author

    Wu, Yan-Xia ; Gu, Guo-chang ; Wang, Ke-Hui

  • Author_Institution
    Harbin Eng. Univ., Harbin
  • Volume
    5
  • fYear
    2007
  • fDate
    19-22 Aug. 2007
  • Firstpage
    2986
  • Lastpage
    2989
  • Abstract
    Satellite-borne embedded systems require the properties of low-powered and reliability in the spatial radiation environment. The control flow checking is an effective way for the running systems to prevent the broken-down caused by Single Event Upsets. Traditional software control flow checking uses a great deal of branch instructions to detect errors, thus brings great overhead in power dissipation. In this paper, a partition method of basic block is suggested. In this partition method, branch instructions are reduced greatly, while the high error detection coverage remain ensure. The simulated results show that compared with the traditional Control Flow Checking by Software Signatures(CFCSS) control flow checking algorithm, the Improved algorithm can reduce total branch instructions by over 10%, reduce the power dissipation by nearly 9%, without decreasing the error detection coverage.
  • Keywords
    artificial satellites; embedded systems; power aware computing; program compilers; program diagnostics; branch instruction; error detection; partition method; power-aware software control flow checking; satellite-borne embedded system; single event upset; software compilation; software signature; spatial radiation environment; Control systems; Cybernetics; Error correction; Fault tolerance; Hardware; Machine learning; Partitioning algorithms; Power dissipation; Power system reliability; Software algorithms; Base node; Control flow checking; Control flow graph; Low power dissipation;
  • fLanguage
    English
  • Publisher
    ieee
  • Conference_Titel
    Machine Learning and Cybernetics, 2007 International Conference on
  • Conference_Location
    Hong Kong
  • Print_ISBN
    978-1-4244-0973-0
  • Electronic_ISBN
    978-1-4244-0973-0
  • Type

    conf

  • DOI
    10.1109/ICMLC.2007.4370659
  • Filename
    4370659