An adaptive large neighbourhood search for designing transparent optical core network

Abstract

In this paper, we focus on designing a core optical network. Given a set of Metro-Core (MC) nodes, the problem is to select a set of links such that the maximum distance between any pair of nodes is bounded. The primary objective is to minimise the total length of the links followed by a secondary objective which is to minimise the total length of the paths between the pairs of MC nodes. We present a mixed integer programming (MIP) model for this problem. As the size of the instances of this problem increases, the MIP model does not scale in terms of memory and time. We therefore present an adaptive large neighbourhood search (LNS) algorithm where the idea is to find an initial solution and repeatedly improve it by solving relatively small subproblems. We propose two different ways of selecting subproblems. The experiments are performed using 3 real-world network topologies: Ireland, UK and Italy. Our results demonstrate that the developed approach is suitable for solving large networks and it can compute very high quality solutions which are very close to optimal values.