Nature-inspired reconfiguration of overlay networks

Many different kind of overlay networking technologies have emerged over the past years. The research has focused on the development of technologies for solving the challenges to provide reliable and efficient networks for data delivery. The ability to automatically reconfigure overlay networks is essential for the future of autonomous computer systems. This paper introduces two approaches for the automatic reconfiguration of overlay networks. The first is a Genetic Algorithm approach as proposed in previous literature, and the second is a Particle Swarm Optimization approach. Both approaches are implemented in order to compare and evaluate their response to link failures and complete network failures.