This paper explores the application of genetic algorithms (GAs) in designing networks, particularly focusing on networks with fluctuating demands. The authors investigate how different genetic operators, such as inversion, mutation, and crossover, affect the quality of the optimal network. They examine the trade-off between pleiotropy ( multiple tasks performed by a single component) and redundancy ( multiple components performing the same task), and how these factors influence the cost and reliability of the network. The study uses a graphical user interface (GUI) to facilitate the evolution of the network and allows for easy modification of parameters. The fitness function is defined to minimize cost and maximize reliability, and the authors introduce a new cost function that better captures the system's connectivity. The results show that the crossover operator significantly improves the convergence rate and the exploration of the solution space compared to mutation alone. The study also evaluates the impact of varying link failure probabilities and population sizes on the convergence time and network parameters. Overall, the research provides insights into optimizing network design using evolutionary computation techniques.This paper explores the application of genetic algorithms (GAs) in designing networks, particularly focusing on networks with fluctuating demands. The authors investigate how different genetic operators, such as inversion, mutation, and crossover, affect the quality of the optimal network. They examine the trade-off between pleiotropy ( multiple tasks performed by a single component) and redundancy ( multiple components performing the same task), and how these factors influence the cost and reliability of the network. The study uses a graphical user interface (GUI) to facilitate the evolution of the network and allows for easy modification of parameters. The fitness function is defined to minimize cost and maximize reliability, and the authors introduce a new cost function that better captures the system's connectivity. The results show that the crossover operator significantly improves the convergence rate and the exploration of the solution space compared to mutation alone. The study also evaluates the impact of varying link failure probabilities and population sizes on the convergence time and network parameters. Overall, the research provides insights into optimizing network design using evolutionary computation techniques.