A Deterministic Approach for Diagnosis Test Generation

The main purpose of the diagnosis test pattern generation is to give the enough information for fault diagnosis during the manufacturing time. The most of the test generation methods are simulation based and have probabilistic nature. In this paper I propose the deterministic approach, which guarantees fault distinguishability. Algorithm is single stuck-at fault based. I perform some optimization on algorithms procedures, so that this algorithm obtains polynomial complexity and return the minimal test set.