On the use of fault dominance in n-detection test generation

The size of an n-detection test set increases approximately linearly with n. This increase in size may be too fast when an upper bound on test set size must be satisfied. We propose a method for obtaining a more gradual increase in the sizes of n detection test sets, while still ensuring that every additional test would be useful in improving the test set quality. The method is based on the use of fault dominance relations to identify a small subset of faults whose numbers of detections are likely to have a high impact on the defect coverage of the test set