Genetic Test Generation for Single and Multiple Crosstalk Faults

Abstract

The main models of crosstalk faults are considered: induced positive and negative pulses, induced delays. A method for modeling cross-faults in a multivalued alphabet was developed. On this basis, a genetic algorithm has been developed for test generation for single cross-faults. The problem of test generation for faults of the type induced delay is considered. A genetic algorithm for test generation for induced delay faults with many aggressor lines is proposed. It was shown for combinational digital circuits application of genetic algorithms and multivalued event-driven simulation in multivalued alphabet allows effectively solve test generation problem for crosstalk faults and gives high fault coverage.