Abstract: This paper introduces a method for evolving combinational digital circuits using a genetic algorithm guided by the NK fitness landscape model. The circuit design process is formulated as a search across a fitness landscape, where circuits with the highest fitness (100%) are selected from each generation. The evolutionary approach employs Cartesian Genetic Programming (CGP) for circuit representation and evaluates candidate circuits using two distinct fitness functions to identify the most efficient designs. By integrating the NK landscape framework, the research analyzes the structure and properties of fitness landscapes—such as ruggedness and neutrality—and their impact on the evolutionary search. The results demonstrate that the proposed methodology can generate optimized and regular digital circuits, confirming the effectiveness of evolutionary techniques in circuit design. The comparison of fitness functions further highlights the benefits of selecting appropriate evaluation metrics for guiding evolution toward optimal circuit solutions.

Keywords: CGP Programming, Genetic Algorithm, Landscape Fitness, NK landscapes, Fitness functions, Digital circuits

Pages: 88-94