An optimization model for MPLS networks

Telecommunication networks are currently evolving towards a universal and convergent network, known as next generation network (NGN), capable of flowing multiservice traffic (voice, data and video) over the same IP based infrastructure. Due to the specific characteristics of each type of traffic, the network must treat them in a differentiated way in order to provide the quality of service (QoS) demanded by users. This fact may stimulate service providers to improve network planning techniques to adequately provide network resources. In this context, multiprotocol label switching (MPLS) plays an important role, contributing for doing an efficient traffic engineering (TE) in order to provide the QoS and to optimize resource allocation. For this purpose, it is possible to apply mixed integer mathematical programming (optimization) techniques to model routing problem minimizing implantation, operation and maintenance network costs. In this paper, we present an optimization model applied to the MPLS network planning, which assign paths (label switched paths-LSP) based on capacity and network architecture constraints, considering LSP survivability. The model is consolidated in two case studies, applying them to plan a hypothetical MPLS network. The main characteristics of the MPLS concerning to the network planning approach are presented.