Pedro Sousa*, Miguel Rocha*, Miguel Rio**, Paulo Cortez***
|
Universidade do Minho Departamento de Informática* Departamento de Sistemas de Informação*** P-4710-057 Braga, Portugal
Tel.: +351 253 604430 |
UCL (University College London)** Department of Computer Science London WC1E 6BT United Kingdom
Tel: +44 20 7679 7214 |
This paper presents a contribution for the development of network management tools able to automatically provide QoS aware routing configurations. In this perspective, a traffic engineering framework able to provide near-optimal OSPF configurations is presented. The devised solution takes into account the multiconstrained QoS demands of a given network domain in order to improve the quality of the OSPF weight setting process. To pursue this objective, this work resorts to Evolutionary Algorithms, which provide OSPF configurations based on a bi-objective function. The proof of concept of the proposed optimization framework resorts to a wide range of simulation studies, clearly showing the effectiveness of the devised mechanisms.