Test Generation 171 In certain cases, it may not be possible for the ATPG to deduce that all vectors with some primary input set to a given logic value definitely do not qualify to be solution vectors. However, it may be able to make an intelligent guess that input vectors with primary input # i set to some specific logic value are more likely to lead to a solution. In such a case, the ATPG would make a decision on primary input # i . Because the decision may actually be wrong, the ATPG may eventually have to alter its decision, trying the vectors that have the opposite Boolean value on primary input # i . The process of making decisions and reversing decisions will result in a decision tree . Each node in the decision tree represents a decision variable. If only two choices are possible for each decision variable, then the decision tree is a binary tree. However, there may be cases where multiple choices are possible in a general search tree. Figure 4.7 shows an example decision tree. While this figure only allows decisions
This is the end of the preview. Sign up
access the rest of the document.