Beyond the Byzantine Generals: Unexpected Behavior and Bridging Fault Diagnosis

Physical defects cause behaviors unmodeled by even the best fault simulators, which complicates predictive diagnosis. This paper reports on a diagnosis procedure that uses modified composite signatures constructed from single stuck-at information combined with a lexicographic matching and ranking algorithm. The diagnosis procedure is used to perform highquality bridging fault diagnosis for more than 400,000 diagnostic experiments involving dropping or adding behaviors from the simulations of faulty circuits. 1 Introduction Accurate fault diagnosis is an integral part of failure analysis. The purpose of fault diagnosis is to specify a part of the chip to examine in order to determine the cause of failure. Traditional fault diagnosis compares the observed faulty behavior to behaviors predicted by a chosen fault model and fault simulator. As identified by Aitken and Maxwell [4], the problem has been approached in two ways: via a simple fault model (almost always the single stuck-at model..