Improving evolvable hardware by applying the speciation technique

Evolvable hardware (EHW) has recently become a highly attractive topic of study because it offers a way of adapting hardware to a given embedded environment. However, it is not easy to evolve hardware efficiently and effectively, so many challenges continue to exist when trying to solve problems. In this paper, we propose a method that uses the speciation technique to enable diverse circuits to evolve efficiently by the process of one-step evolution. As a result of studying the landscape contained in the EHW example, we have found complicated spaces contain many peaks that can lead to deceptions when using the evolving process, and the speciation technique profits from the evolution of EHW. We also studied that the speciated hardware ensemble might be a good candidate for more complex and rigorous function. In the experiments, we applied the fitness sharing method as the speciation technique, and obtained diverse hardware modules, then ascertained the efficiency of these structures. We also show that several useful extra functions and better overall performance can be obtained by analyzing diverse circuits with the speciation technique.