Subscribe

Issue No.10 - October (2001 vol.50)

pp: 1095-1102

DOI Bookmark: http://doi.ieeecomputersociety.org/10.1109/12.956094

ABSTRACT

<p><b>Abstract</b>—A new diagnosis algorithm for square grids is introduced. The algorithm always provides correct diagnosis if the number of faulty processors is below <tmath>$T$</tmath>, a bound with <tmath>$T\in\Theta(n^{2/3})$</tmath>, which was derived by worst-case analysis. A more effective tool to validate the diagnosis correctness is the syndrome dependent bound <tmath>$T_\sigma$</tmath>, with <tmath>$T_\sigma\geq T$</tmath>, asserted by the diagnosis algorithm itself for every given diagnosis experiment. Simulation studies provided evidence that the diagnosis is complete or almost complete if the number of faults is below <tmath>$T$</tmath>. The fraction of units which cannot be identified as either faulty or nonfaulty remains relatively small as long as the number of faults is below <tmath>$n/3$</tmath> and, as long as the number of faults is below <tmath>$n/2$</tmath>, the diagnosis is correct with high probability.</p>

INDEX TERMS

System-level diagnosis, PMC model, processor grids, constant-degree diagnosis, diagnosis algorithm.

CITATION

Stefano Chessa, Piero Maestrini, "Correct and Almost Complete Diagnosis of Processor Grids",

*IEEE Transactions on Computers*, vol.50, no. 10, pp. 1095-1102, October 2001, doi:10.1109/12.956094