Automatic test program generation:: A case study

A simulation-based methodology different from pure or biased pseudorandom methods was proposed. The approach calls for the generation of a set of test programs and optimization using the feedback information from a simulator able to evaluate them with respect to a given coverage metric. The use of such a method can dramatically help designers and engineers: Instead of checking massive random simulations for differences with respect to the correct model, validation experts can let the automatic test case generator work for the proper time and eventually examine the test set it produces.