Elastic Network Design and Adaptive Flow Placement in Software Defined Networks

Abstract

Large scale telecommunication network deployment and its management requires huge Operational Expenditure (OPEX) and Capital Expenditure (CAPEX) from the network operator. Network virtualization, Software Defined Networks (SDN) concepts and Cloud principles enable innovative approaches for introducing elasticity in the network for reducing OPEX. The activation and de-activation of network elements and their efficient interconnection is known as Network Design enabling up- and downscaling of the parts of the entire network on demand based on the traffic situation. Adaptive Flow Placement through SDN controller functionalities introduces novel mechanisms to influence service data flows within the network actively with the goal to distribute the network load equally, increase reliability and reduce cost through deactivation of unused network parts. This paper introduces a novel cost saving approach for on demand elastic Network Design and active Flow Placement in SDN environments. The underlying Network Design problem is being solved using a State of the Art Mixed Integer Programming (MIP) Solver and a sophisticated MIP formulation. A summary concludes this paper.