Diagnostic Fault Equivalence Identification

A significant problem with current diagnostic test generation techniques is the time spent in identifying diagnostic equivalences amongst fault pairs. First, fault pair distance analysis is introduced to characterize diagnostically equivalent fault pairs. It is shown that diagnostically equivalent fault pairs are usually “close” to each other and influence the same set of primary outputs. The distance analysis provides a motivation for local circuit transformations and structural analysis to identify equivalences in combinational circuits rapidly. Our results establish a connection between redundant faults and a specific class of diagnostically equivalent fault pairs. Structural analysis is then used to identify equivalences between fault pairs. All the techniques developed are integrated into an equivalence identification tool that is used as a front end to a diagnostic test generation system. Experimental results are presented on the ISCAS85 and ISCAS89 benchmark circuits to demonstrate the efficiency of the techniques.