Synthesizing Clustered, Secured, and Hierarchical Networks through Genetic Algorithms

We have formulated and combined three problems: clustering, firewall placement and network hierarchy into one optimization problem, where the objective function is to minimize the total design cost of the synthesized network, while maximizing its security and scalability. Due to the computational complexity of the combined three problems, we have developed a custom-made intelligent algorithm based on the genetic algorithms (GA) to search the design space for good solutions. We have conducted five experiments with different mutation rates for three different network scenarios comprising of 50, 70, and 100 clients respectively. Our results show that the custom-made GA has converged and optimized the cost of network by 60% from its initial design.