Flexible cross layer optimization for fixed and mobile broadband telecommunication networks and beyond

As the telecommunication world moves towards a data-only network environment, signaling, voice and other data are similarly transported as Internet Protocol packets. New requirements, challenges and opportunities are bound to this transition and in uence telecommunication architectures accordingly. In this time in which the Internet in general, and telecommunication networks in particular, have entered critical infrastructures and systems, it is of high importance to guarantee e cient and exible data transport. A certain level of Quality-of-Service (QoS) for critical services is crucial even during overload situations in the access and core network, as these two are the bottlenecks in the network. However, the current telecommunication architecture is rigid and static, which o ers very limited exibility and adaptability. Several concepts on clean slate as well as evolutionary approaches have been proposed and de ned in order to cope with these new challenges and requirements. One of these approaches is the Cross Layer Optimization paradigm. This concept omits the strict separation and isolation of the Application-, Controland NetworkLayers as it enables interaction and fosters Cross Layer Optimization among them. One indicator underlying this trend is the programmability of network functions, which emerges clearly during the telecommunication network evolution towards the Future Internet. The concept is regarded as one solution for service control in future mobile core networks. However, no standardized approach for Cross Layer signaling nor optimizations in between the individual layers have been standardized at the time this thesis was written. The main objective of this thesis is the design, implementation and evaluation of a Cross Layer Optimization concept on telecommunication networks. A major emphasis is given to the de nition of a theoretical model and its practical realization through the implementation of a Cross Layer network resource optimization system for telecommunication systems. The key questions answered through this thesis are: in which way can the Cross Layer Optimization paradigm be applied on telecommunication networks; which new requirements arise; which of the required functionalities cannot be covered through existing solutions, what other conceptual approaches already exist and nally whether such a new concept is viable. The work presented in this thesis and its contributions can be summarized in four parts: First, a review of related work, a requirement analysis and a gap analysis were performed. Second, challenges, limitations, opportunities and design aspects for specifying an optimization model between application and network layer were formulated. Third, a conceptual model Generic Adaptive Resource Control (GARC) was speci ed and its prototypical implementation was realized. Fourth, the theoretical and practical thesis contributions was validated and evaluated.