A novel approach to the analysis of VLSI device test programs

Being short of the formal theory and methodology of program debugging, the aim of the investigation in this paper is to create an approach to support analysis and debugging of VLSI device test programs. In research into the previous work, we find that there are two restrictions in program debugging: computational complexity and diagnostic precision, which greatly influence the effect of program debugging. We apply Fuzzy Mathematics, especially Fuzzy Comprehensive Evaluation, to researches on fuzzy knowledge representation, fuzzy relation, and fuzzy inference, etc. in the test programs analysis. In the paper, we analyze the construction of the test program, present the concept of test entities, and build the evaluation space of the test entities, etc. Lastly we address an algorithm, with an example, of the selection of the test entities to be debugged.