Signal probability calculations using partial functional manipulation

Abstract

Signal probability calculations are necessary to determine the random pattern testability of logic circuits. Determination of random pattern testability is necessary for considering the use of weighted or unweighted linear feedback shift registers (LFSRs) as an appropriate testing method. This paper presents an algorithm to accurately and efficiently (both in space and time) calculate signal probabilities (sometimes called syndrome analysis) within digital logic networks. It has the advantage that it uses a new method for signal probability calculations which is typically both fast and accurate, and which has already efficiently produced results for all the ISCAS combinational circuits.<>