Knowledge-driven and component based network architecture for applications with high quality demands 

Multimedia services over broadband DSL access and aggregation networks such as Broadcast TV (BCTV), Network Based Personal Video Recording and Video on Demand (VoD) have gained a lot of popularity in the last few years. For end users, these services introduce new possibilities such as interactivity and higher video quality. For service operators, they offer increased revenue in their Triple Play offer. At the same time, multimedia services have stringent quality requirements: in order to avoid visual distortions they often require a substantial amount of bandwidth and tolerate no packet loss and only small amounts of jitter. Operators who want to maximize their revenue try to manage the service quality as perceived by the end user, commonly described as the Quality of Experience (QoE). This management is further complicated by the heterogeneity of today's access and aggregation networks. Autonomic networking manages this growing complexity by adding intelligence inside network nodes and network management applications. The autonomic communications paradigm aspires to build a self-governing network which allows administrators to focus more on the optimization of the network through higher level goals instead of performing time consuming manual management tasks. In this research, an autonomic communications architecture is targeted that is able to maximize the QoE in multimedia access networks. In the first term of this research, three aspects of an autonomic communications system were studied: (1) the use of enabling components such as monitoring and QoE optimizers, (2) the reasoning on the knowledge to deploy new QoE optimizers and (3) the exchange of context between autonomic managers.