While majority is focusing on the sensors, communication and network aspects of IoT systems. This thesis investigates into energy efficient aspect of scheduling messages by the things/ objects in IoT. Things or objects are clustered into IoT subgroups where each subgroup has a message broker that delivers the messages originating from members to the ultimate receiver of sensed data Le.sink. This work proposes message scheduling algorithms which improves the overall efficiency of the IoT systems. It considers the network layer routing aspect to keep the energy depletion low to provide a more optimal solution by applying certain degree of cross-layer design methodology. It proves the effectiveness and efficiency both in terms of service response time and energy consumption via simulation results. Furthermore, as messages have different priorities, an algorithm is designed considering this aspects as well. Messages are classified into high priority (HP) and best effort (BE) by the corresponding Quality of Service (QoS) aware scheduling in each IoT subgroup to differentiate emergency messages from non-mission critical messages. An energy efficient backup nodes selection algorithm is presented if any node becomes non responsive due to some internal error or external reason. This approach finds an energy efficient optimal level of replacement nodes for IoT systems, while keeping in view energy of sensor devices and the cost of deployment of the backup nodes.

The purpose of this study is to attain information on women's use of the Internet by focusing on the ways it has empowered them. Tehran is selected as a field site because nearly one-fifth of the urban population in Iran lives in Tehran. Moreover, the Access to the Internet, particularly in this capital city, has developed enormously in the recent years. The main questions around which I have focused this enquiry are: "Has the Internet made a difference in the lives of women in the city of Tehran? Do they feel empowered by the Internet? This research is based on semi structured interviews which consist of a written list of ten questions. The research sample consists of 40 participants between the ages 13-45 years, whom I have selected through the snowballing sampling technique. The data for this study was collected during the summer 2010. The aim of this research is to attain information on women's use of the Internet and their perception about empowerment, mainly focusing on how much the use of the Internet contributes to their empowerment and in which ways does this occur. The data reveals that 80 per cent of the interviewed women acknowledge that the Internet has a significant impact on their personal and professional life. This group also believes that the Internet has opened up a new space for them. Their empowerment zones were focused on the aspects of individual (physical, psychological, and intellectual wellbeing), political, economic, social and cultural empowerment. However, 20 per cent of the participants stress that the Internet 'may' have made their life easier but has not brought about a big difference to their lives practically. The reason of this disagreement according to their views is related to the major obstacles they encounter in their easy access to the Internet.

RSS (really simple syndication, rich site summary or RDF site summary) is a dialect of XML that provides a method of syndicating on-line content, where postings consist of frequently updated news items, blog entries and multimedia. RSS feeds, produced by organisations or individuals, are often aggregated, and delivered to users for consumption via readers. The semi-structured format of RSS also allows the delivery/exchange of machine-readable content between different platforms and systems. Articles on web pages frequently include icons that represent social media services which facilitate social data. Amongst these, RSS feeds deliver data which is typically presented in the journalistic style of headline, story and snapshot(s). Consequently, applications and academic research have employed RSS on this basis. Therefore, within the context of social media, the question arises: can the social function, i.e. utility, of RSS be enhanced by producing from it data which is actionable and effective? This thesis is based upon the hypothesis that the fluctuations in the keyword frequencies present in RSS can be mined to produce actionable and effective data, to enhance the technology's social utility. To this end, we present a series of laboratory-based case studies which demonstrate two novel and logically consistent RSS-mining paradigms. Our first paradigm allows users to define mining rules to mine data from feeds. The second paradigm employs a semi-automated classification of feeds and correlates this with sentiment. We visualise the outputs produced by the case studies for these paradigms, where they can benefit users in real-world scenarios, varying from statistics and trend analysis to mining financial and sporting data. The contributions of this thesis to web engineering and text mining are the demonstration of the proof of concept of our paradigms, through the integration of an array of open-source, third-party products into a coherent and innovative, alpha-version prototype software implemented in a Java JSP/servlet-based web application architecture.

Technology has proven to be a pivotal factor that helps to drive organizational effectiveness, because it helps to automate processes, improve speed of data processing, enhance organizational performance and ensure ease of exchange of information across the organization. One of such emerging technologies that is fast shaping the organizational landscape is cloud computing; and this research focused on the relevance of this technology to the oil and gas sector, the benefits, risks and analysing how the implications of adopting this technology can be addressed. Also, it has been established that adopting cloud based solutions is related to outsourcing, and this implies that another firm (cloud provider) will manage the function or process that is migrated to the cloud. Therefore, this research is important within the Oil and gas sector because the success of adopting this technology in the sector is based on how these companies can manage the diverse inter-firm relationships with cloud providers. To draw new insights, relational view approach was applied to gain the understanding of how managing these relationships can ensure successful adoption of the technology because existing literatures lack in-depth discussion on how organizations can manage multiple cloud deployments (especially in the Oil sector). Furthermore, the research aimed at understanding the current state of cloud deployment in this sector, determine the most suitable deployment model(s), and strategies for effective adoption, taking cognizance of the complexity of the operations within the industry. This research focused on the Nigerian Oil industry as the research setting for gathering data for this research. This is important because this sector is the major source of revenue for Government in Nigeria and any approach that helps in achieving performance improvement will automatically translate into more benefits for the country and other stakeholders in the sector. In view of this, this research was conducted qualitatively by conducting interviews for participants within the Nigerian Oil and gas sector. These participants cut across Oil exploration and production companies, Cloud providers, and Government/Regulator agencies. The data gathering was focused on understanding the extent of adoption within the sector, current usage of this technology (if any), implications, how the transitioning is being coordinated and risk and strategies currently being deployed. The findings of this research show a low level of adoption of cloud solutions, despite the moderate level of awareness of this technology. This low level of adoption was premised on some factors ranging from the conservative culture in the industry, infrastructure challenges (peculiar with Nigeria), suitability of cloud solutions for specific core oil and gas operations, security and confidentiality concerns, lack of suitable corporate strategies, regulatory concerns, just to mention a few. This research analysed these findings critically and proposed the need for corporate strategy, effective stakeholder collaboration and management and how to adopt suitable governance framework. Furthermore, the need for National cloud policy and how to address infrastructure challenges which are inhibiting cloud adoption in the research area (Nigeria) were also addressed.

In this thesis, we study the problem of computing approximate equilibria in several classes of games. In particular, we study approximate Nash equilibria and approximate well-supported Nash equilibria in polymatrix and bimatrix games and approximate equilibria in Lipschitz games, penalty games and biased games. We construct algorithms for computing approximate equilibria that beat the cur- rent best algorithms for these problems. In Chapter 3, we present a distributed method to compute approximate Nash equilibria in bimatrix games. In contrast to previous approaches that analyze the two payoff matrices at the same time (for example, by solving a single LP that combines the two players’ payoffs), our algorithm first solves two independent LPs, each of which is derived from one of the two payoff matrices, and then computes an approximate Nash equilibrium using only limited communication between the players. In Chapter 4, we present an algorithm that, for every δ in the range 0 < δ ≤ 0.5, finds a (0.5+δ)-Nash equilibrium of a polymatrix game in time polynomial in the input size and 1 . Note that our approximation guarantee does not depend on δ the number of players, a property that was not previously known to be achievable for polymatrix games, and still cannot be achieved for general strategic-form games. In Chapter 5, we present an approximation-preserving reduction from the problem of computing an approximate Bayesian Nash equilibrium (ε-BNE) for a two-player Bayesian game to the problem of computing an ε-NE of a polymatrix game and thus show that the algorithm of Chapter 4 can be applied to two-player Bayesian games. Furthermore, we provide a simple polynomial-time algorithm for computing a 0.5-BNE. In Chapter 5, we study games with non-linear utility functions for the players. Our key insight is that Lipschitz continuity of the utility function allows us to provide algorithms for finding approximate equilibria in these games. We begin by studying Lipschitz games, which encompass, for example, all concave games with Lipschitz continuous payoff functions. We provide an efficient algorithm for computing approximate equilibria in these games. Then we turn our attention to penalty games, which encompass biased games and games in which players take risk into account. Here we show that if the penalty function is Lipschitz continuous, then we can provide a quasi-polynomial time approximation scheme. Finally, we study distance biased games, where we present simple strongly poly- nomial time algorithms for finding best responses in L1, L2, and L∞ biased games, and then use these algorithms to provide strongly polynomial algorithms that find 2/3, 5/7, and 2/3 approximations for these norms, respectively.

This thesis is in the field of Computer Science. More precisely, its main research themes are in the applied part of the field Cloud Computing. The main focus in this work is on monitoring of cloud systems in a distributed fashion. This work is a natural continuation of previous studies on discovering the symptoms malicious behaviours in cloud systems. Our line of research is based on efficient discovery of the symptoms of threats. This challenge is met through the design and analysis of new algorithms carrying out this job. Several algorithms are studied. First, a simplified version of previously studied Mobility algorithm is proposed. The new algorithm is named Reduce-Max algorithm. This algorithm is analysed on eight different data sets. Then two modifications of Reduce-Max algorithm are considered. The first one is called Randomised-Local Reduction and the second one is Deterministic-Centralised Reduction. Further, the algorithms are tested under different models of symptoms appearance. The work continues with studies of Reduce-Max and its two modifications in hierarchical systems, which concludes in the design of a new algorithm, called Random-Start-Round-Robin. Finally, this thesis concludes with work on balancing Mobility Algorithm. An integral part of my PhD work are experiments of proposed algorithms where the emphasis is on proper modeling of monitoring of cloud systems. Further discussion is based on the results of these experiments reflected in the final conclusions.

Wireless sensor networks (WSN) within Internet of Things (IoT) have the potential to address the growing detection and classification requirements among many surveillance applications. RF sensing techniques are the next generation technologies which offer distinct advantages over traditional passive means of sensing such as acoustic and seismic which are used for surveillance and target detection applications of WSN. RF sensing based WSN within IoT detect the presence of designated targets by transmitting RF signals into the sensing environment and observing the reflected echoes. In this thesis, an RF sensing based target detection architecture for surveillance applications of WSN has been proposed to detect the presence of stationary targets within the sensing environment. With multiple sensing nodes operating simultaneously within the sensing region, diversity among the sensing nodes in the choice of transmit waveforms is required. Existing multiple access techniques to accommodate multiple sensing nodes within the sensing environment are not suitable for RF sensing based WSN. In this thesis, a diversity in the choice of the transmit waveforms has been proposed and transmit waveforms which are suitable for RF sensing based WSN have been discussed. A criterion have been defined to quantify the ease of detecting the signal and energy efficiency of the signal based on which ease of detection index and energy efficiency index respectively have been generated. The waveform selection criterion proposed in this thesis takes the WSN sensing conditions into account and identifies the optimum transmit waveform within the available choices of transmit waveforms based on their respective ease of detection and energy efficiency indexes. A target detector analyses the received RF signals to make a decision regarding the existence or absence of targets within the sensing region. Existing target detectors which are discussed in the context of WSN do not take the factors such as interference and nature of the sensing environment into account. Depending on the nature of the sensing environment, in this thesis the sensing environments are classified as homogeneous and heterogeneous sensing environments. Within homogeneous sensing environments the presence of interference from the neighbouring sensing nodes is assumed. A target detector has been proposed for WSN within homogeneous sensing environments which can reliably detect the presence of targets. Within heterogeneous sensing environments the presence of clutter and interfering waveforms is assumed. A target detector has been proposed for WSN within heterogeneous sensing environments to detect targets in the presence of clutter and interfering waveforms. A clutter estimation technique has been proposed to assist the proposed target detector to achieve increased target detection reliability in the presence of clutter. A combination of compressive and two-step target detection architectures has been proposed to reduce the transmission costs. Finally, a 2-stage target detection architecture has been proposed to reduce the computational complexity of the proposed target detection architecture.

Cloud computing offers an innovative business model to enterprise IT services consumption and delivery. However, vendor lock-in is recognised as being a major barrier to the adoption of cloud computing, due to lack of standardisation. So far, current solutions and efforts tackling the vendor lock-in problem have been confined to/or are predominantly technology-oriented. Limited studies exist to analyse and highlight the complexity of vendor lock-in problem existing in the cloud environment. Consequently, customers are unaware of proprietary standards which inhibit interoperability and portability of applications when taking services from vendors. The complexity of the service offerings makes it imperative for businesses to use a clear and well understood decision process to procure, migrate and/or discontinue cloud services. To date, the expertise and technological solutions to simplify such transition and facilitate good decision making to avoid lock-in risks in the cloud are limited. Besides, little research investigations have been carried out to provide a cloud migration decision framework to assist enterprises to avoid lock-in risks when implementing cloud-based Software-as-a-Service (SaaS) solutions within existing environments. Such decision framework is important to reduce complexity and variations in implementation patterns on the cloud provider side, while at the same time minimizing potential switching cost for enterprises by resolving integration issues with existing IT infrastructures. Thus, the purpose of this thesis is to propose a decision framework to mitigate vendor lock-in risks in cloud (SaaS) migration. The framework follows a systematic literature review and analysis to present research findings containing factual and objective information, and business requirements for vendor-neutral interoperable cloud services, and/or when making architectural decisions for secure cloud migration and integration. The underlying research procedure for this thesis investigation consists of a survey based on qualitative and quantitative approaches conducted to identify the main risk factors that give rise to cloud computing lock-in situations. Epistemologically, the research design consists of two distinct phases. In phase 1, qualitative data were collected using open-ended interviews with IT practitioners to explore the business-related issues of vendor lock-in affecting cloud adoption. Whereas the goal of phase 2 was to identify and evaluate the risks and opportunities of lock-in which affect stakeholders’ decision-making about migrating to cloud-based solutions. In synthesis, the survey analysis and the framework proposed by this research (through its step-by-step approach), provides guidance on how enterprises can avoid being locked to individual cloud service providers. This reduces the risk of dependency on a cloud provider for service provision, especially if data portability, as the most fundamental aspect, is not enabled. Moreover, it also ensures appropriate pre-planning and due diligence so that the correct cloud service provider(s) with the most acceptable risks to vendor lock-in is chosen, and that the impact on the business is properly understood (upfront), managed (iteratively), and controlled (periodically). Each decision step within the framework prepares the way for the subsequent step, which supports a company to gather the correct information to make a right decision before proceeding to the next step. The reason for such an approach is to support an organisation with its planning and adaptation of the services to suit the business requirements and objectives. Furthermore, several strategies are proposed on how to avoid and mitigate lock-in risks when migrating to cloud computing. The strategies relate to contract, selection of vendors that support standardised formats and protocols regarding data structures and APIs, negotiating cloud service agreements (SLA) accordingly as well as developing awareness of commonalities and dependencies among cloud-based solutions. The implementation of proposed strategies and supporting framework has a great potential to reduce the risks of vendor lock-in.

The field of Cloud Computing is relatively new branch of Information Technology in which various services are devolved from a centralised local location to a de-centralized remote Intranet/Internet environment. It has recently experienced rapid growth and acceptance with academia and industry, presenting new challenges worthy of fundamental research. Some of the largest challenges today revolve around achieving higher levels of sustainability and infrastructure performance. This work investigates an optimised and novel approach to an Autonomous Virtual Server Management System in a ‘Cloud Computing’ environment through designing and building an Autonomous Management Distributed System (AMDS). The AMDS helps reduce hardware power consumption through autonomously moving virtual servers around a network to balance out hardware loads, as well as being easily configurable and extendable, made possible by its software infrastructure. Through use of an internally configured Cloud Computing test -bed rig, the AMDS makes use of several physically and logically defined networks to communicate with all devices that are a part of the cloud infrastructure. Once connected, the AMDS monitors these devices and issues optimisation commands accordingly. Experimental results show an overall power consumption reduction of up to 8%, which in a typical datacentre of several thousand servers translates into a significant cost reduction. This work also presents an initial design, along with proof-of-concept implementation as an AMDS module, of a Botnet heuristic detection algorithm. Experimental results show an overall malicious data packet detection rate of 52%, a significant figure for only 5000 data samples analysed by the module. Another strength is that this design allows an abstract software model to be constructed, which can then be implemented using a multitude of programming languages. This research shows how the carbon footprint of a Cloud Computing datacentre can be reduced and reveals a significant impact on issues of sustainability with respect to both energy efficiency and economic viability. It also shows how datacentre security can be enhanced by detecting Botnet activity and preventing the disruption of day-to-day operations through a highly scalable, flexible, and autonomous software implementation.

Cloud computing technology offers flexible, pay-per-use and convenient access to a pool of services and virtualised computer resources using internet connection. Despite these benefits, the adoption of cloud computing by Small and Medium Enterprises (SMEs) is still slow due to (perceived) security and privacy issues. Recent studies concluded that such issues could result in issues of trust for both adopters and potential adopters of cloud computing. While security and privacy issues are actively being researched in the area of cloud computing, there is little published research regarding the aspect of trust between the clients (SMEs) and their Cloud Service Providers (CSPs). The main focus of this study was to investigate the role of trust and other factors involved in the adoption and usage of cloud computing by SMEs. By combining the variables introduced by Diffusion of Innovation Theory (Rogers, 2003), Technology Organisation Environment Framework (Tornatzky and Fleishchner, 1990) and the Integrative Model of Organisational Trust (Mayer, et al., 1995), a conceptual model was produced. This model was tested empirically through an online survey of 269 participants consisting UK SMEs. Using the statistical software ‘SPSS’, the description of each variable was presented. The reliability of multi scale items was assessed using Cronbach’s alpha. Factor analysis was carried out to reduce the dimensions of items used for further analysis (regression). Then an ordinal regression analysis was done to examine the relationship between variables. It was found that an increase in the challenges of cloud computing decreases its chances of adoption. Also, an increase in the knowledge level of cloud computing was found to increase the chances of adopting cloud computing. On the other hand, trust in service provider was found to have a negative effect on the perceived usefulness of cloud computing. This is because majority of the respondents revealed that cloud computing is very useful but indicated total disagreement of trust in their CSPs. This is not an attractive finding for the CSPs. Therefore, the recommendations provided will enable to CSPs to increase trust in order to encourage the continuous use of cloud computing by adopters and also encourage the uptake of cloud computing by potential adopters.