Context aware cross-layer architecture for real-time wireless local area networking

The exponential growth of the wireless Internet demands staying connected while on the move. It is necessary that wireless technologies support the growing demand for wireless real-time services. Challenges to efficient real-time communications over wireless networks arise due to the inherent limitations of wireless networks, resource demands of real-time applications and the inefficiency of the layered protocol stack in enabling dynamic adaptations. At the same time, context awareness is an essential and integral component of next generation pervasive computing. Hence, there is an increasing realisation of the necessity of a generic cross-layer context-aware framework for efficient real-time wireless communication to overcome limitations of the existing layered stack. Even though individual aspects of cross-layer interactions have been studied in the literature, a holistic view of cross-layer interactions from the viewpoint of context awareness is not available. Further, as pervasive computing is driving the demand for anywhere, any time and from any device services, it is necessity to enable underpinning mobility management in an efficient and adaptive manner. The research presented in this dissertation proposes a cohesive and holistic cross-layer framework called (CA)2RW-Com which can support cross-layer context awareness while preserving modularity of the layered protocol stack and hence ensuring interoperability. The significant components of context awareness including context acquisition, context representation, cross-layer context exchange and context management are defined in the proposed framework. The dissertation proposes context acquisition through minimum changes to the existing protocol stack hence ensures uninterrupted operation of noncross-layer protocols and applications. Modelling and representing cross-layer context and adaptive situations in a unifying and generic way is discussed. A modular cross-layer signalling mechanism is proposed for context exchange within the framework. Further, the flexibility of the framework is addressed through its ability to support various joint viii multi-layer and single-layer adaptations. Enabling entity-executed and system-executed single-layer adaptations in the framework is also discussed. The dissertation proposes three adaptations in the mobility management domain. The first isjoint multi-layer adaptation. In this adaptation, application, network and data link layers jointly support mobility management for heterogeneous networks. The second and third adaptations are single-layer mobility management solutions. Context-Aware Adaptive Multi-homed Mobile IP is an extension to Multi-homed Mobile IP which is proposed as a single-layer entity-executed adaptation that the (CA)2RW-Com framework supports. Adaptive Performance Metric is a single-layer system-executed adaptation proposed for Access Point selection of IEEE 802.11 networks. We have modelled and validated supporting joint multi-layer mobility management proposed in the (CA)2RW-Com framework. Further, we have enabled, implemented and evaluated the single-layer mobility management solutions we propose in a simulation environment. Moreover, the dissertation discusses the prototype implementation of the (CA)2RW-Com framework and communication specification. The dissertation also presents the operational statistics of the prototype. In summary, this research presents a cohesive and holistic approach for defining a generic context-aware cross-layerframework for wirelesslocal area networks. Further, it proposes and validates cross-layer context-aware adaptive mobility management approaches. To date, the results of this research have been published in ten peer-reviewed publications and one journal article is currently under review.