214VLSI Test Principles and Architecturesfitness metric can mislead the GA to arrive at a suboptimal solution, or even nosolution at all. For instance, a population whose individuals’ fitnesses are similarwill not allow the selection process to identify more highly fit individuals to act asparents. Furthermore, without a metric, the individuals may become indistinguish-able even when they really are distinguishable. For example, if the fitness functionis simply a binary function, where an individual’s fitness is equal to one if the targetfault is excited and zero otherwise, this will result in many individuals with fitnessequal to zero if they do not excite the target fault. It is clear, however, that someindividuals may be closer to exciting the target fault than others. However, theaforementioned binary fitness function would prevent the GA from distinguishingthose more fit individuals from the less fit ones.
This is the end of the preview.
access the rest of the document.