No description available.

Ad hoc wireless networking offers mobile computer users the prospect of trading with others in their vicinity anywhere anytime. This thesis explores the potential for developing such trading applications. A notable difficulty in designing their security services is being unable to use trusted parties. No one can be guaranteed present in each ad hoc wireless network session. A side benefit is that their costs don't have to be paid for. A reference model is defined for ad hoc m-commerce and a threat model is for- mulated of its security vulnerabilities. They are used to elicit security objectives and requirements for such trading systems. Possible countermeasures to address the threats are critically analysed and used to design security services to mitigate them. They include a self-organised P2P identity support scheme using PGP cer- tificates; a distributed reputation system backed by sanctions; a group membership service based on membership vouchers, quorate decisions by some group members and partial membership lists; and a security warning scheme. Security analysis of the schemes shows that they can mitigate the threats to an adequate degree to meet the trading system's security objectives and requirements if users take due care when trading within it. Formal verification of the system shows that it satisfies certain safety properties.

The emerging technologies of harvesting the energy from the environment surrounding the application have recently attracted intensive attention of design automation researchers.Due to the universal presence of vibrations on machines, among the different mechanisms available to obtain electrical power from ambient energy, the vibration basedharvesters have been the subject of particularly extensive development. A vibration-based (kinetic) harvester simply converts vibrations in the environment surrounding a wireless sensor node into electrical energy. This enables the wireless sensor node to be placed anywhere in the environment with no need for access to facilitate battery replacement. The basic structure of vibration-based energy harvester is composed of multi-domain components, it contains electrical, mechanical, and magnetic in the case of electromagnetic harvester. In addition, many design parameters from different domains need to be optimised in a holistic manner (i.e. treating all the system components as a connected unit); all these requirements besides the traditional approaches of optimisation, complicate the hardware description language for analog and mixed syste (HDL-AMS) simulation and makes central processor unit (CPU) takes prohibitive time, even with today's multi-physics simulation tools. This research develops, a novel optimisation technique, which enables effcient optimisation and design exploration for such a complex system and reduce CPU computation time for optimisation purposes. The proposed methodology accelerates the optimisation by approximately two orders of magnitude due to the utilisation of the design of experiment (DoE) approach and response surface modelling (RSM). The contributions of this research can be summarised as follows: Firstly, a novel, response surface based design space exploration approach to energy harvester powered systems has been developed. The proposed technique enables designers to gain insight into the details of design parameters trade-offs and quantifies each design parameter effect on performance indicators via the response surface mathematical model. The method has been applied to a linear micro-electromagnetic cantilevered harvester. Secondly, a novel, fast performance optimisation technique for a wireless sensor node powered by a tunable kinetic energy harvester has been developed. The result of applying this technique reduces the total CPU optimisation time by two orders of magnitude compared with the classical approach, i.e. through multiple full simulations. Thirdly, a software tool set has been created, based on MATLAB and VHDL-AMS, for fast, multi-dimensional design space exploration and optimisation of a kinetic harvester

A Mobile Ad Hoc Network (MANET) is more prone to security threats than other wired and wireless networks because of the distributed nature of the network. Conventional MANET routing protocols assume that all nodes cooperate without maliciously disrupting the operation of the protocol and do not provide defence against attackers. Blackhole and flooding attacks have a dramatic negative impact while grayhole and selfish attacks have a little negative impact on the performance of MANET routing protocols. Malicious nodes or misbehaviour actions detection in the network is an important task to maintain the proper routing protocol operation. Current solutions cannot guarantee the true classification of nodes because the cooperative nature of the MANETs which leads to false exclusions of innocent nodes and/or good classification of malicious nodes. The thesis introduces a new concept of Self- Protocol Trustiness (SPT) to discover malicious nodes with a very high trustiness ratio of a node classification. Designing and implementing new mechanisms that can resist flooding and blackhole attacks which have high negative impacts on the performance of these reactive protocols is the main objective of the thesis. The design of these mechanisms is based on SPT concept to ensure the high trustiness ratio of node classification. In addition, they neither incorporate the use of cryptographic algorithms nor depend on routing packet formats which make these solutions robust and reliable, and simplify their implementations in different MANET reactive protocols. Anti-Flooding (AF) mechanism is designed to resist flooding attacks which relies on locally applied timers and thresholds to classify nodes as malicious. Although AF mechanism succeeded in discovering malicious nodes within a small time, it has a number of thresholds that enable attacker to subvert the algorithm and cannot guarantee that the excluded nodes are genuine malicious nodes which was the motivation to develop this algorithm. On the other hand, Flooding Attack Resisting Mechanism (FARM) is designed to close the security gaps and overcome the drawbacks of AF mechanism. It succeeded in detecting and excluding more than 80% of flooding nodes within the simulation time with a very high trustiness ratio. Anti-Blackhole (AB) mechanism is designed to resist blackhole attacks and relies on a single threshold. The algorithm guarantees 100% exclusion of blackhole nodes and does not exclude any innocent node that may forward a reply packet. Although AB mechanism succeeded in discovering malicious nodes within a small time, the only suggested threshold enables an attacker to subvert the algorithm which was the motivation to develop it. On the other hand, Blackhole Resisting Mechanism (BRM) has the main advantages of AB mechanism while it is designed to close the security gaps and overcome the drawbacks of AB mechanism. It succeeded in detecting and excluding the vast majority of blackhole nodes within the simulation time.

Network bandwidth and server capacity are gradually becoming overloaded due to the high demand and rapid evolution of high quality multimedia services over the Internet. Internet Protocol Television (IPTV) is among the multimedia services that demand more of network and server resources, especially with the emergence of Mobile IPTV. It is imperative for the service providers to maintain good quality management services in order to satisfy their clients. To guarantee the required quality of service (QoS) and quality of experience (QoE) in IPTV, the server must have the required capacity and resources to serve all the clients’ requests. The flexibility of IPTV services which provide users with the ability to stream multimedia content at anytime and anywhere they want, makes the demand for video-on-demand (VoD) services higher. However, the server bandwidth capacity is limited, and as such the numerous requests from the clients may exhaust all the available bandwidth depending on the number of requests at a given time, which may lead to the poor QoS and QoE. In this research, a new algorithm called Intelligent Routing Algorithm for Mobile IPTV (IRA-MIPTV) is proposed. The algorithm combined features and advantages of Internet Protocol (IP), Mobile Ad hoc Network (MANET) characteristics and Content Delivery Network (CDN) based network architecture to improve on the QoS and QoE in mobile IPTV. The proposed algorithm is aimed at reducing total dependency on the server by the mobile nodes. The algorithm intelligently learns the best server or client to serve an incoming service request depending on the available server capacity and the number of requests received at a point in time. The routing decision is made by the Designated Server (DS) that selects and reroutes a request to the most appropriate server or client. The novelty of this research work can simply be identified as the designing, developing and evaluating an Intelligent Routing Algorithm for mobile ITPV (IRA-MIPTV) that intelligently learns and select a reflective server or client to serve a particular service request on behalf of the server during high service demand. The selection depends on the server available bandwidth, load and proximity. The proposed algorithm also dynamically adjusts to server failure by assigning the role of designated server to the backup server and re-elect another backup server to guarantee service delivery at all times. To validate the effectiveness of the proposed algorithm, different simulation tests were conducted using OPNET/Riverbed Modeler 18.0. A typical IPTV network, where packets are delivered over IP, and the proposed algorithm were modelled and incorporated into the Modeler. For the study to reflect more on real situations, live video programme was streamed using VLC media player. The packet’s size and packet inter-arrival time data were collected and used in the simulation’s environment. After conducting a series of simulation tests, the results showed that the proposed algorithm outperforms the normal IPTV system in server load reduction, high throughput and low amount of end-to-end delay, as well as adaptability and robustness. The results also showed that the efficiency of the proposed algorithm increases as the number of clients increase. It also confirmed that the algorithm reduces the server overload during high service request periods by using clients to serve some of the incoming service requests on behalf of the server. The algorithm produced low server and network load, low end-to-end delay, high throughput, adaptability and robustness.

The central cause of strain on today's IP network infrastructure steins from the industries striving to extend the network to carry voice and other advanced services that it was not designed for. This thesis develops the reasons why more sophisticated mechanisms are required for ensuring that the best-effort infrastructure becomes both predictable and controllable, if the current IP infrastructure is to replace circuit switched networks. The main theme of the thesis are the IP traffic engineering (IPTE) traffic engineering algorithms that were developed by the author. The goal of the proposed approach is to compute a set of Open Shortest Path First (OSPF) link weights to balance network load, while honouring bandwidth, hop count and propagation delay constraints of the traffic. The proposed solution is built on classical OSPF routing and additionally uses the multi topology concepts that were recently proposed as an extension to OSPF. Using Multi Topology OSPF (mt-OSPF), multiple routing planes can be implemented. Each Multi Topology (MT) routing plane is assigned its own link metrics and thus traffic can be routed independently. Given a traffic demand matrix and the network topology, the algorithm computes a set of link weights using a search heuristic. The optimisation is cost function based, so that individual constraints can be taken into account per routing plane. Using this approach, constraint information remains in the "offline" IPTE algorithms and thus no extra constraint awareness is required at layer 3. Since recent Cisco implementations of mt-OSPF and Multi Topology IS-IS (M-ISIS) provide the required multi topology routing support, no major changes at the router level are required for the approach to be realised. Extensive simulations presented in this thesis show that IPTE has the potential to provide the differentiated routing and load balancing that it was designed for. In addition the simulations show that load is balanced more evenly across the network than with standard shortest path routing on inverse capacity link weights. A network management system is discussed that acts as the binding element for all enabling components. A simplified architecture for such a management system is presented and it is discussed how several traffic engineering mechanisms could coexist with IPTE on one network infrastructure. The use of the authors IPTE approach in the context of this management system is discussed in depth: Dangers associated with potential network disruption caused by frequent link weight modifications are analysed from both intra- and inter-domain point of view. A strategy for solving the problem of disruptions caused by link weight changes is presented by using several network planes to migrate traffic, rather than causing disruptions to a "live" plane through link weight modification and resulting OSPF routing table updates. The strategy is simulated, showing that any detrimental effects of the transition are avoidable. However, the work is ongoing and the results presented are indicative. Overall the thesis presents an approach for traffic engineering using traditional routing techniques, to (1) preserve the original intrinsic advantages of the IP design and (2) help ready IP networking for the much more restricting requirements of the future mutiservice network. It presents algorithms, simulations and frameworks showing how the work fits into the current IP networking world.

No description available.

No description available.

We have categorised the characteristics and the content of pervasive computing environments (PCEs), and demonstrated why a non-dynamic approach to knowledge conceptualisation in PCEs does not fulfil the expectations we may have from them. Consequently, we have proposed a formalised computational model, the FCM, for knowledge representation and reasoning in PCEs which, secures the delivery of situation and domain specific services to their users. The proposed model is a user centric model, materialised as a software engineering solution, which uses the computations generated from the FCM, stores them within software architectural components, which in turn can be deployed using modern software technologies. The model has also been inspired by the Semantic Web (SW) vision and provision of SW technologies. Therefore, the FCM creates a semantically rich situation-specific PCE based on SWRL-enabled OWL ontologies that allows reasoning about the situation in a PCE and delivers situation specific service. The proposed FCM model has been illustrated through the example of remote patient monitoring in the healthcare domain. Numerous software applications generated from the FCM have been deployed using Integrated Development Environments and OWL-API.

Web applications incorporate important business assets and offer a convenient way for businesses to promote their services through the internet. Many of these web applica- tions have evolved from simple HTML pages to complex applications that have a high maintenance cost. This is due to the inherent characteristics of web applications, to the fast internet evolution and to the pressing market which imposes short development cycles and frequent modifications. In order to control the maintenance cost, quantita- tive metrics and models for predicting web applications’ maintainability must be used. Maintainability metrics and models can be useful for predicting maintenance cost, risky components and can help in assessing and choosing between different software artifacts. Since, web applications are different from traditional software systems, models and met- rics for traditional systems can not be applied with confidence to web applications. Web applications have special features such as hypertext structure, dynamic code generation and heterogenousity that can not be captured by traditional and object-oriented metrics. This research explores empirically the relationships between new UML design met- rics based on Conallen’s extension for web applications and maintainability. UML web design metrics are used to gauge whether the maintainability of a system can be im- proved by comparing and correlating the results with different measures of maintain- ability. We studied the relationship between our UML metrics and the following main- tainability measures: Understandability Time (the time spent on understanding the soft- ware artifact in order to complete the questionnaire), Modifiability Time(the time spent on identifying places for modification and making those modifications on the software artifact), LOC (absolute net value of the total number of lines added and deleted for com- ponents in a class diagram), and nRev (total number of revisions for components in a class diagram). Our results gave an indication that there is a possibility for a relationship to exist between our metrics and modifiability time. However, the results did not show statistical significance on the effect of the metrics on understandability time. Our results showed that there is a relationship between our metrics and LOC(Lines of Code). We found that the following metrics NAssoc, NClientScriptsComp, NServerScriptsComp, and CoupEntropy explained the effort measured by LOC(Lines of Code). We found that NC, and CoupEntropy metrics explained the effort measured by nRev(Number of Revi- sions). Our results give a first indication of the usefulness of the UML design metrics, they show that there is a reasonable chance that useful prediction models can be built from early UML design metrics.