Comparative evaluation of genetic algorithm and memetic algorithm for optimal topology design of computer networks

Abstract

Network topology design problem can be formulated as a combinatorial optimization problem. In this paper, a comparative evaluation of genetic algorithm and memetic algorithm is presented with a view to investigating their performance for network design problem. Results from various computation experiments reveal that MA is able to produce better results in term of quality but at a higher computation time. In addition, for a problem instance of small size such as 10-node network design problem GA can produce better solution but at a higher computational time.

Keywords: Memetic Algorithm, Genetic Algorithm, local search, optimization, network