Test-vector prediction of M-testable iterative arrays

Abstract

In this paper, a new M-testability approach is presented as a further development of C-testability concept (limited to regular iterative arrays) to cope with more general arrays (i.e. of non-identical cells). M-testability is proposed based on a developed Classified-Level Approach (CLA), applied to the interconnected cells as a first step toward their test-vector prediction, under the assumption of at most one-faulty cell. The test-vector prediction is conducted on each of the defined cell classes. Using an elaborated Variable Testability Measure (VTM), the number of test vectors are predicted for each cell, then for the entire array. Applicable regardless of the fault type, M-tetability is experimented on various iterative arrays and generated results are debated.