Automatic SLA management in SLA-aware architecture

While proposals for improvements to be introduced in the network in order to support a real differentiation of services follow each other, network operators seem to face these requirements by making overprovisioning on their IP infrastructures. However, no mechanism is offered to request different end-to-end services at different prices. At the moment the interconnections among IP networks are established by means of SLAs which require high manual overhead and a high associated cost which limits their wide-spread use. This causes scarce providers' revenues and a non-optimised resource allocation of resources inside the network. Hence, it seems now clear that such agreements cannot be managed statically and on a long term timescale, and that the issue of an automatic SLA management has to be addressed. This paper introduces a new entity based on emerging Elastic Networks paradigm for dynamic management of SLAs, and describes the interaction of such an entity with an architecture capable to provide SLA enforcement on top of QoS-enabled networks.