A hybrid distributed test generation method using deterministic and genetic algorithms

Test generation is a highly complex and time-consuming task. In this work, we present a distributed method for combinational test generation. The method is based on a hybrid approach that combines both deterministic and genetic approaches. The deterministic phase is based on the D-algorithm and generates an initial set of test vectors that are evolved in the genetic phase in order to achieve high fault coverage in a short time. The algorithm is parallelized based on a cluster of workstations using the message passing interface (MPI) library. Several benchmark circuits were attempted, and favorable results comparisons are reported.