A Cartesian Genetic Programming Approach for evolving Optimal Digital Circuits
S. Kazarlis, John A. Kalomiros, V. Kalaitzis · Journal of Engineering Science and Technology Review · 2016
In this work, a Cartesian Genetic Programming (CGP) method is presented for the evolutionary synthesis of optimal digital circuits.In this evolutionary method, circuits are represented as directed graphs in the form of an MxN Cartesian grid.Evolved circuits are qualified using a fast custom emulator of digital circuits that was built to serve as a fitness function for the algorithm.The performance of the CGP algorithm is tested on six different digital circuits, used as benchmarks.The evolution results prove the ability of the CGP scheme to find optimal solutions with significant probabilities.Moreover, the CGP algorithm is able to produce unconventional solutions for known circuits.