Mobile Ad Hoc Networks (MANETs) have emerged as a new paradigm in communication networks, enabling pervasive computing and ubiquitous communication environments. Their main characteristic is the mobility that nodes exhibit, with the network topology changing potentially rapidly and unpredictably. Traditional wireless networks require some form of fixed network infrastructure and centralised administration for their operation. On the contrary, MANETs are spontaneously formed, with individual nodes responsible for dynamically discovering other nodes they can communicate with. The key benefits of MANETs such as the lack of centralised administration, tether-less computing capabilities and community- based short-term network establishment are hindered by the management difficulties caused by the flexible infrastructure that underlies these benefits. There is a need for frameworks that can support the self-management of MANETs according to predefined goals or policies. We assert that such a highly dynamic environment can potentially benefit from context information that will drive its self-management, resulting in a degree of autonomy. High-level management rules expressed as policies - to achieve extensibility and flexibility - can guide the MANET configuration, triggered by context information gathered by every node and disseminated across the MANET for network-wide understanding to be established. This closed-loop adaptive management can thus lead to self-configuration, self-optimisation, and hence autonomy. This thesis addresses all critical aspects of the design, implementation and deployment of a context-aware framework to achieve self-management of MANETs. Context awareness achieved through context modelling and efficient monitoring is the foundation of this thesis. It feeds the management scheme that is built on the requirements of a MANET-oriented organisational model. Based on dynamic context information, the management scheme enforces through a programmable configuration platform changes in the MANET nodes so as to conform to the high-level management decisions. These decisions may in turn result in new context information being generated and thus new management decisions to be taken. Consequently a closed loop self-management cycle is generated, which can lead to self-optimisation, self-protection and self-healing of MANETs, subject to the appropriate network policies having been introduced into the system. The proposed context-aware framework for the self-management of MANETs is generic enough to support management changes based on any type of monitored context information, In order to validate the proposed framework, we have selected a particular application case study that relates to an adaptive routing protocol strategy. The relative mobility of the nodes in a MANET is monitored and relevant information triggers routing protocol selection and deployment. The deployed routing protocol is the appropriate one for the monitored conditions. Evaluation of this work was done using analytical modelling, simulation and also testbed experimentation. Obviously our framework is not restricted to dynamic routing strategies but allows for dynamic strategies of diverse nature to be deployed on MANETs in a generic fashion based on any type of context information. Keywords: context-awareness, context modelling, mobile ad hoc networks, network management.

Multimodal interaction is one of the taxonomies for Human-Computer Interaction (HCI). With the introduction of multimodal interactions, input/output information is becoming associated with the different human senses so that information can be presented in the most efficient and natural way. However, in mobile communication, a number of restrictions are still remnant. Mostly, these restrictions are caused by limitations of a mobile terminal's user interfaces. This thesis introduces an architectural framework to facilitate multimodal interaction in a virtual- device environment. The framework developed is called the Multi Interface-Device Binding (MID-B) system. MID-B provides the functions and features to overcome the drawbacks of classic multimodal interaction. In the classical sense, multimodality uses a strategy that simultaneously utilises several modalities generally offered on a 'single' device. In contrast, the MID-B's mechanisms take multimodality out of the single-device scenario. In MID-B, a 'controller-device' is aware of the availability of various devices in the vicinity, each of which may host one or more user interfaces (modalities).- The capabilities of those co-located devices, together with the context in which-the user acts, are exploited to dynamically customise the interface services available. MID-B binds these devices into a virtual device to exploit their- individual user interfaces (modalities) in a combined way. This thesis describes the MID-B architecture and its mechanisms to collect the context information of 'devices' and 'users'. The thesis presents the methodologies to exploit that context information to dynamically adapt user interfaces.

The work in this thesis describes a system that allocates the end-to-end bandwidth, in a meshed communications network, to set up calls. The solution makes use of market-based software agents that compete in a number of decentralised market-places in order to buy and sell bandwidth. The network is partitioned into regions, where each has a separate market server from where the bandwidth resources are allocated. Agents perform a distributed search with decentralised markets in order to allocate routes for calls. The approach relies on a resource reservation and commit mechanism in the network. Results show that for a typical network given a light load, the system sets up 95% of calls and is comparable in performance to an instantaneous global search, which assumes complete knowledge of resource availability and sets up 97% of calls at the same load. As the network load is increased to a typical intermediate level, the system gives an average call success rate which is 76% of that given by the instantaneous global search. Furthermore, under all network loads, the system performs significantly better than a random strategy. We also consider issues associated with scalability and robustness in our multi-agent system. Specifically, an empirical evaluation was carried out to assess the system performance under a variety of design configurations in order to provide an insight into network deployment issues. This took place using networks that were fixed in size but which were scaled up with respect to the number of regions and, hence, market servers. We investigated two main measures: the average call success rate and average message load per market server, as the number of markets were increased in the network. The same investigations were performed in the presence of single market failures. For both the failure and non-failure cases, a trade-off was found between the two measures to find an optimum number of regions to deploy in the network. In general, results showed that in the absence of failures, fewer regions gave a higher average call success rate and more regions meant that, on average, market servers received less messages per market server, when searching for resources across a single regional route. With a single market failure, we saw that when more regions were deployed, in general, the average call success rate was higher up to a certain point and the number of messages received per market server was lower, when also using a single regional route. In the final part of this thesis, we make a change to part of our original algorithm in order to try and improve upon the system performance. To overcome the need for receiving fewer messages per market server and having quicker call set up times, we allow buyer agents to bid for resource bundles that lead to several boundary nodes in a region, in a single auction, rather than the possibility of using multiple auctions. Thus, we introduce a multi-destination buyer strategy. By constraining the amount of resource searching in this way, we can see that there are performance trade-offs where call set up times decrease, average call success rates increase in general, but that the system allocates a smaller proportion of longer distance calls than when using the original strategy. We also show that, in general, it is better to search across a single regional route instead of searching across multiple regional routes, particularly when the network load is high. However, at much lighter network loads, searching for resources by means of traversing alternative regions can give a better overall average call success rate.

The aim of this research is to develop a model of trust that will endeavour to assure good interactions amongst autonomous software agents in complex, networked environments. In this context, we identify the following as key characteristics. Firstly, such environments are open, meaning that agents are free to enter and exit the system at their will, so an agent cannot be aware of all of its interaction partners. Furthermore, there is a possibility that these interaction partners may be malicious or colluding agents. Secondly, the openness and dynamism of these environments means agents will need to interact with other agents, with which they have had no past experience. Even in this context, an agent must be able to accurately assess the trustworthiness of another. Thirdly, the distributed and heterogeneous nature of these systems influences any model or application developed for such environments. Specifically, this often requires models and applications to be decentralised. Lastly, many of the interactions that occur between agents in such systems are in the context of a virtual organisation (VO). Here VOs are viewed as collections of agents belonging to different organisations, in which each agent has a specific problem solving capability which when combined provides a particular service to meet the requirements of an end user. Now, VOs are social structures, and the presence of certain inter-agent relationships may influence the behaviour of certain members. For this reason it is important to consider not only personal experiences with an individual to determine its behaviour, but to also examine the social relationships that it has with other agents. Against this background, we have developed TRAVOS (A Trust and Reputation Model for Agent-Based Virtual Organisations) which focuses, in particular, on providing a measure of trust for an agent to place in an interaction partner. This measure of trust is calculated by considering the past experiences between the agent and its interaction partner. In instances when there is no personal experience, the model substitutes past experience with reputation information gathered from other agents in the society or from special reputation broker agents. Reputation is gathered in a way that filters out biased or false opinions. In addition to this, the model is constrained by issues of scalability and decentralisation. Furthermore, by extending TRAVOS we developed a set of mechanisms (TRAVOS-R) related to learning and exploiting the social relationships present in VO-rich environments. More specifically, TRAVOS-R presents a novel approach to learning the type of relationship present between two agents, and uses this knowledge to adjust the opinions obtained from one agent about the other. The TRAVOS models have been tested empirically and have significantly outperformed other similar models. Moreover, to further evaluate the applicability of our approach a realistic system evaluation was also carried out, which involved applying our models in an industrial application of agent-based VOs. In undertaking this research, we have shown that trust is a key component of networked systems and that a computational trust model can be used by agents in large, dynamic, uncertain and open environments to account for the uncertainty inherent in their social decision-making processes. More specifically, we have shown that by using personal experience, opinions from others, and knowledge of social relationships, an agent is able to arrive at a more accurate trust value, and, as a consequence, that it can interact in a more effective manner.

After the explosion of Internet and World Wide Web use in the 1990s, the Internet Protocol has become the de facto standard for networked multimedia communications, sidestepping theoretically superior technologies like Frame Relay and ATM. This fact, combined with the higher efficiency of packet switching over circuit switching, has resulted in Voice over IP (VoIP) networks being increasingly preferred to the General Switched Telephone Network (GSTN), a trend that continues to accelerate in the 21st century. However, inherent technical difficulties related to the predominantly "best-effort" nature of IP, have restricted VoIP to the inside of corporate intranets or to the backbone of carrier networks, and no universal "IP Switched Telephone Network" (ISTN) exists to replace the GSTN. This thesis investigates the applicability of the ISTN concept, on the assumption that the establishment of such an infrastructure will require enhanced scalability and tighter integration of technologies related to all three areas of end-to-end communication: the LAN (end users), for whom appropriate applications and usage models must be proposed, so that IP Telephony can be widely adopted the MAN (access networks), particularly those interfacing with the GSTN, so that interoperability is secured and the WAN (backbone), where call routing is expected to be a crucial factor in the deployment of a unified, worldwide VoIP infrastructure. Examining the proposals currently used in the above three areas, reveals a number of shortcomings, which render the introduction of an alternative solution a necessity for ISTN to be realised in the near future. Therefore, a new architecture for large-scale IP Telephony is introduced and evaluated in this thesis, using improved designs for three core components, one per area addressed: a user agent for the LAN, a gateway for the MAN and a call routing mechanism for the WAN, which is generally more scalable than existing (and under-researched) ones. Finally, potential enhancements to the proposed architecture are suggested and certain directions for further investigation are identified.

There are very many different ways how context information could be used to make computer systems and applications more user-friendly, flexible, and adaptable. The use of context information is especially important in a mobile environment, where the environment of interaction, execution, and usage needs change rapidly. Areas where increased use of context information can bring added value and where research work is conducted include human computer interaction, adaptable user interfaces, virtual and augmented reality, mobile, ubiquitous, handheld, and wearable computing. The objectives of this thesis were to briefly clarify what context and context-awareness mean and to analyse the network context-awareness in networks through the use of policies. This was done through the development of a Context Policy Based Management Framework (Context Aware Policy Language and Policy Based Management Architecture), which was initially functionally tested at component level, and eventually successfully exercised by implementing two context aware services through the use of context aware policies.

In this thesis, the channel assignment and routing problems have been investigated for both cooperative and competitive Wireless Mesh networks (WMNs). A dynamic and distributed channel assignment scheme has been proposed which generates the network topologies ensuring less interference and better connectivity. The proposed channel assignment scheme is capable of detecting the node failures and mobility in an efficient manner. The channel monitoring module precisely records the quality of bi-directional links in terms of link delays. In addition, a Quality of Service based Multi-Radio Ad-hoc On Demand Distance Vector (QMR-AODV) routing protocol has been devised. QMR-AODV is multi-radio compatible and provides delay guarantees on end-to-end paths. The inherited problem of AODV’s network wide flooding has been solved by selectively forwarding the routing queries on specified interfaces. The QoS based delay routing metric, combined with the selective route request forwarding, reduces the routing overhead from 24% up to 36% and produces 40.4% to 55.89% less network delays for traffic profiles of 10 to 60 flows, respectively. A distributed channel assignment scheme has been proposed for competitive WMNs, where the problem has been investigated by applying the concepts from non-cooperative bargaining Game Theory in two stages. In the first stage of the game, individual nodes of the non-cooperative setup is considered as the unit of analysis, where sufficient and necessary conditions for the existence of Nash Equilibrium (NE) and Negotiation-Proof Nash Equilibrium (N-PNE) have been derived. A distributed algorithm has been presented with perfect information available to the nodes of the network. In the presence of perfect information, each node has the knowledge of interference experience by the channels in its collision domain. The game converges to N-PNE in finite time and the average fairness achieved by all the nodes is greater than 0.79 (79%) as measured through Jain Fairness Index. Since N-PNE and NE are not always a system optimal solutions when considered from the end-nodes prospective, the model is further extended to incorporate non-cooperative end-users bargaining between two end user’s Mesh Access Points (MAPs), where an increase of 10% to 27% in end-to-end throughput is achieved. Furthermore, a non-cooperative game theoretical model is proposed for end-users flow routing in a multi-radio multi-channel WMNs. The end user nodes are selfish and compete for the channel resources across the WMNs backbone, aiming to maximize their own benefit without taking care for the overall system optimization. The end-to-end throughputs achieved by the flows of an end node and interference experienced across the WMNs backbone are considered as the performance parameters in the utility function. Theoretical foundation has been drawn based on the concepts from the Game Theory and necessary conditions for the existence of NE have been extensively derived. A distributed algorithm running on each end node with imperfect information has been implemented to assess the usefulness of the proposed mechanism. The analytical results have proven that a pure strategy Nash Equilibrium exists with the proposed necessary conditions in a game of imperfect information. Based on a distributed algorithm, the game converges to a stable state in finite time. The proposed game theoretical model provides a more reasonable solution with a standard deviation of 2.19Mbps as compared to 3.74Mbps of the random flow routing. Finally, the Price of Anarchy (PoA) of the system is close to one which shows the efficiency of the proposed scheme.