On invariant implication relations for removing partial circuits
Hideyuki Ichihara, Kozo Kinoshita, Seiji Kajihara · Systems and Computers in Japan · 1997
In redundancy removal for combinational circuits using SOCRATES-based test generation, it is required that static learning be repeated whenever a redundant fault is removed, because implications learned for the original circuit may change. In this paper we discuss invariance of implication relations derived in static learning for the modifying circuit by redundancy removal, then propose an efficient redundancy removal procedure in which only invariant implications are used for redundancy identification without repeating static learning. Experimental results for benchmark circuits show that the proposed method is up to 60 times faster than the previous method. © 1997 Scripta Technica, Inc. Syst Comp Jpn, 28(7): 39–47, 1997