Artificial Bee Colony for optimization of cloud-ready and survivable elastic optical networks

Abstract

In this paper we study important and weakly-investigated optimization problem imposed by the fast telecommunication networks development. The problem concerns routing of three types of flows (unicast, anycast, multicast) in survivable Elastic Optical Network (EON). Three types of flows allow to efficiently support cloud-related services while network survivability, provided by dedicated path protection, addresses problem of a single link failure. Since the problem is very challenging, we propose efficient and large-scale optimization method based on the Artificial Bee Colony (ABC) framework. Then, we perform extensive numerical experiments to evaluate ABC performance with respect to the reference methods. The results prove high efficiency of the ABC-based algorithm, which finds very good solutions (average gap to optimal result was less than 10% for all studied cases) and outperforms the reference methods for the majority of problem instances.