Evolution of Secure Networks through Simulation

2 The rapid spread of computer worms and viruses has become a major threat to governments, businesses, educational institutions, and even home computing. In order to combat this threat, a computational procedure based on evolution principles was introduced to create a more secure computer network. A series of computer programs written in Perl was developed and used in conjunction with a Perl package known as the Network Worm Simulator (NWS) to effectively simulate large-scale computer networks and control the evolution. The evolutionary system was run for 100 generations of 10 networks, or 1000 networks, in order to generate the optimized output. The optimized networks showed an average of 50 % improvement on average infection rate. In addition, it was found that network infection rate is inversely proportional to the number of generations simulated, and that distributed network topologies lead to more secure networks. Some validation was given by SSFNet under a global Internet environment. However, more model improvements are needed for realistic implementations