Distributed Ant Algorithm for Inter-Carrier Service Composition

Abstract

The future Internet would benefit from the delivery of high performance services crossing several carriers and bringing new incomes to telco operators. Such services compel carriers to guarantee Quality of Service (QoS) inside their boundaries. As carriers are independently administrated, architectures and processes must be reconsidered so that requirements for autonomy and confidentiality are respected. In this context, enforcing Service Level Agreements (SLAs) between partners cannot be avoided. The following problem has to be solved: given an end-to-end service request and collections of SLAs on each participating carrier, establish an end-to-end contract, which gives an optimal service under reliable QoS guarantees. In previous works, we described distributed processes based on dynamic programming principles. In this article, we depict a distributed adaptation of a meta-heuristic algorithm, known as Ant Colony algorithm. We demonstrate that the performances of this algorithm are competitive to other works.