Title
A Combinatorial Group Testing Method For Fpga Fault Location
Keywords
Combinatorial group testing; Dueling and halving algorithms; Fault-handling
Abstract
Adaptive fault isolation methods based on discrepancy-enabled pairwise comparisons are developed for reconfigurable logic devices. By observing the discrepancy characteristics of multiple Concurrent Error Detection (CED) configurations, fault isolation is realized without requiring additional test vectors or data coding schemes. Hence the reprogrammability of Field Programmable Gate Arrays (FPGAs) is utilized to examine CED alternatives in succession. Results show that for a reprogrammable device with one million resources, where 50% of the resources are used on an average by the target application, fault isolation can be achieved in as few as 28 iterations. The effect of resource utilization, the number of competing candidate solutions, and the number of unit resources are analyzed and the performance of a halving-based algorithm for fault isolation are quantified.
Publication Date
12-1-2006
Publication Title
Proceedings of the Seventh IASTED International Conference on Advances in Computer Science and Technology
Number of Pages
55-60
Document Type
Article; Proceedings Paper
Personal Identifier
scopus
Copyright Status
Unknown
Socpus ID
33847209633 (Scopus)
Source API URL
https://api.elsevier.com/content/abstract/scopus_id/33847209633
STARS Citation
Sharma, Carthik A. and DeMara, Ronald F., "A Combinatorial Group Testing Method For Fpga Fault Location" (2006). Scopus Export 2000s. 8110.
https://stars.library.ucf.edu/scopus2000/8110