A tool for random test generation targeting high diagnostic resolution

The main objective of this work is to develop a method and to create a tool for random test pattern generation for digital circuits, capable of achieving a high diagnostic resolution. We propose a measure for evaluating the diagnostic resolution of a given test set. Three methods were investigated using different criterions for selecting test patterns from the given packages of random vectors. The chosen criterion targets the diagnostic resolution as high as possible at the minimum test length and the maximum fault coverage. Experimental research demonstrated that the proposed tool provides the test engineer with better diagnostic resolutions compared to other comparable test generation methods.