Abstract
The authors describe a fault simulator for synchronous sequential circuits given in gate-level. The simulator uses the multiple observation time testing approach to identify faults detected by a given test sequence, that cannot be identified as detected using a conventional fault simulator. It is shown that a test sequence that does not detect a given fault may still be effective in identifying that the circuit is faulty. This situation can only be identified by the multiple observation time approach. Based on this observation, one can propose a new definition of fault coverage, that is more effective than the conventional definition in retaining test sequences (or test vectors) that are effective in identifying faulty circuits. Experimental results are presented to support the effectiveness of multiple observation time fault simulation and of the new fault coverage definition.

This publication has 7 references indexed in Scilit: