• DocumentCode
    2303317
  • Title

    Evolutionary based techniques for fault tolerant field programmable gate arrays

  • Author

    Larchev, Gregory V. ; Lohn, Jason D.

  • Author_Institution
    NASA Ames Res. Center, Moffett Field, CA
  • fYear
    0
  • fDate
    0-0 0
  • Lastpage
    321
  • Abstract
    The use of SRAM-based field programmable gate arrays (FPGAs) is becoming more and more prevalent in space applications. Commercial-grade FPGAs are potentially susceptible to permanently debilitating single-event latchups (SELs). Repair methods based on evolutionary algorithms may be applied to FPGA circuits to enable successful fault recovery. This paper presents the experimental results of applying such methods to repair four commonly used circuits (quadrature decoder, 3-by-3-bit multiplier, 3-by-3-bit adder, 4-to-7 decoder) into which a number of simulated faults has been introduced. The results suggest that evolutionary repair techniques can improve the process of fault recovery when used instead of, or as a supplement to triple modular redundancy (TMR), which is currently the predominant method for mitigating FPGA faults
  • Keywords
    circuit optimisation; circuit reliability; evolutionary computation; fault tolerance; field programmable gate arrays; radiation effects; space vehicle electronics; FPGA fault mitigation; SRAM-based field programmable gate arrays; evolutionary algorithms; evolutionary repair techniques; fault recovery; fault tolerant field programmable gate arrays; single-event latchups; triple modular redundancy; Adders; Circuit faults; Circuit simulation; Decoding; Evolutionary computation; Fault tolerance; Field programmable gate arrays; Manufacturing; NASA; Redundancy;
  • fLanguage
    English
  • Publisher
    ieee
  • Conference_Titel
    Space Mission Challenges for Information Technology, 2006. SMC-IT 2006. Second IEEE International Conference on
  • Conference_Location
    Pasadena, CA
  • Print_ISBN
    0-7695-2644-6
  • Type

    conf

  • DOI
    10.1109/SMC-IT.2006.30
  • Filename
    1659571