• DocumentCode
    3432611
  • Title

    Some analyses on signature analysis with distortionless data compression

  • Author

    Aoyama, Kenji ; Iwadare, Yoshihiro

  • Author_Institution
    Dept. of Inf. Eng., Nagoya Univ., Japan
  • fYear
    1992
  • fDate
    16-20 Nov 1992
  • Firstpage
    1088
  • Abstract
    This paper shows that from the point of view of aliasing errors and the expected values of the wasting test cost caused by misjudging faulty circuits to be fault-free, there are instances where signature analysis with no (circuit output) compression(SA-NC) is preferable to signature analysis with single MISR(SA-MISR) or multiple input shift registers. It is proposed to apply distortionless sequence compression to the circuit output of SA-NC, which preserves the property of all error detection and yet results in come memory element reduction. Huffman coding and modified Ziv-Lempel coding are promising candidates of distortionless compression when the probabilities of occurrence of input sequences are known and unknown respectively. Simulations are made on two examples of circuits under test, a 9-input and 4-output priority encoder, and a 14-input and 8-output arithmetic logic unit. It is shown that in some instances, Huffman and modified Ziv-Lempel coding result in some 50% reduction of the circuit output length
  • Keywords
    Huffman codes; data compression; error detection codes; integrated circuit testing; large scale integration; logic testing; Huffman coding; aliasing errors; arithmetic logic unit; circuit output length; distortionless sequence compression; memory element reduction; modified Ziv-Lempel coding; priority encoder; signature analysis; wasting test cost; Arithmetic; Circuit analysis; Circuit faults; Circuit simulation; Circuit testing; Costs; Data analysis; Huffman coding; Logic testing; Shift registers;
  • fLanguage
    English
  • Publisher
    ieee
  • Conference_Titel
    Singapore ICCS/ISITA '92. 'Communications on the Move'
  • Print_ISBN
    0-7803-0803-4
  • Type

    conf

  • DOI
    10.1109/ICCS.1992.255094
  • Filename
    255094