Exploiting Behavioral Information in Gate-Level ATPG

This paper aims at broadening the scope of hierarchical ATPG to the behavioral-level. A main problem is identified, namely the mismatch of timing models between the behavioral- and gate-levels. As a main contribution of this paper, a theoretical analysis of this problem led to the definition of a novel concept, that of dominated patterns, that captures the needed link between the levels. Some metrics are defined, taken from the software realm, that allow generation of test patterns at the behavioral-level. To validate the concept correctness, different ATPG systems are presented, and experimental results show an improvement in the test quality thanks to the exploitation of behavioral-level information.