A collection of mobile nodes can form a multi-hop radio network with a dynamic topology and without the need for any infrastructure such as base stations or wired network. Such a Mobile Ad Hoc Networks (MANETs) maintain their structure and connectivity in a decentralised and distributed fashion. Each mobile node acts as both a router for other nodes traffic, as well as a source of traffic of its own In this thesis we develop and present a new hybrid routing protocol called Multipath Distance Vector Zone Routing Protocol, which is referred to as MDVZRP. In MDVZRP we assume that all the routes in the routing table are active and usable at any time, unless the node received or discovered a broken link. There is no need to periodically update the routing tables, therefore reducing the periodic update messages and hence reducing the control traffic in the entire network. The protocol guarantees loop freedom and alternative disjoint paths. Routes are immediately available within each routing zone. For destinations outside the zone, MDVZRP employs a route discovery technique known as routing information on demand. Once the node is informed by either the MAC layer or itself that it should discover the non- reachable nodes, MDVZRP adopts a new technique. First, we discuss the Ad Hoc networks and routing in general, then the motivation of MDVZRP regarding the nodes‟ flat view, and the selection and acquisition of multipath getting and selection. Furthermore, we describe the stages of MDVZRP and the protocol routing process with examples. The performance of MDVZRP is then evaluated to determine its operating parameters, and also to investigate its performance in a range of different scenarios. Finally, MDVZRP is compared with DSDV and AODV ordinary routing protocols (standard) delivering CBR traffic. Simulation results show that MDVZRP gives a better performance than DSDV in all circumstances, it is also better than AODV in most of the scenarios, especially at low mobility.

This thesis examines how Software Defined Network (SDN) and Network Virtualization (NV) technologies can make 5G and beyond mobile networks more flexible, scalable and programmable to support the performance demands of the emerging heterogeneous applications. In this direction, concepts like mobile network slicing, multi-tenancy, and multi-connectivity have been investigated and their performance is analyzed. The SDN paradigm is used to enable flexible resource allocation to the end users, improve network resource utilization and avoid or rapidly solve the network congestion problems. The proposed network architectures are 3rd Generation Partnership Project (3GPP) standards compliant and integrate Open Network Foundation (ONF) SDN specifications to ensure seamless interoperability between different standards and backward/forward compatibility. Novel mechanisms and algorithms to efficiently manage the resources of evolving 5G Time-Division Duplex (TDD) networks in a flexible manner are introduced. These mechanisms enable formation of virtual cells on-demand which allows diverse resource utilization from multiple eNBs to the users. Within the scope of this thesis, SDN-based frameworks to enhance the QoE of end user applications considering Time Division-Long Term Evolution (TD-LTE) small cells have also been developed and network resource sharing scenarios with Frequency-Division Duplex (FDD)/TDD coexistence has been studied. In addition, this thesis also proposes and investigates a novel service-oriented network slicing concept for evolving 5G TDD networks which involve traffic prediction mechanisms and includes user mobility. An analytical model is also introduced that formulates the network slice resource allocation as a weighted optimization problem. The evaluations of the proposed solutions are performed using 3GPP standard compliant simulation settings. The proposed solutions have been compared with the state-of-the art schemes and the performance gains offered by the proposed solutions have been demonstrated. Performance is evaluated considering metrics such as throughput, delay, network resource utilization etc. The Mean Opinion Score (MOS) metric is used for evaluating the Quality of Experience (QoE) for end-user applications. With the help of SDN-based network management algorithms investigated in this work, it is shown how 5G+ networks can be managed efficiently, while at the same time provide enhanced flexibility and programmability to improve the performance of diverse applications and services delivered over the network to the end users.

This thesis is concerned with the design and analysis of new Bayesian network based mobility algorithms for mobile Wireless Sensor Networks (WSNs). The hypothesis for the work presented herein is that incorporating Artificial Intelligence (Al) at the level of the sensor nodes will improve their performance (coverage, connectivity and lifetime) and result in fault tolerance capabilities, in the face of uncertainty associated with incomplete information regarding the network. Two types of mobility strategy are presented and investigated. Firstly, a new gazing mobility strategy is presented which is biologically inspired from herbivores grazing pastures. As part of the latter strategy, and instead of deploying a large number of static sensor nodes to cover a region of interest, a smaller number of mobile nodes are deployed which migrate around the region to achieve coverage over time. To enable the performance evaluation of this strategy a new coverage measure called Coverage Against Time was created. A new decentralised Bayesian network based grazing mobility algorithm called BNGRAZ is presented which uses evidence derived from neighbouring nodes to predict the probability of performance (coverage and connectivity) changes associated with moving in a particular direction. Evidence is also obtained from a new Coverage Approximation (CA) algorithm which enables each sensor node to approximate the WSN coverage in order to determine areas in need of servicing. The performance of BNGRAZ is compared to a fixed path mobility technique, Random Waypoint (RWP) mobility model, and a new Grazing Reference Point Group Mobility (GRPGM) algorithm developed as part of this work. Secondly, a self-healing strategy which physically relocates sensor nodes to repair coverage holes, due to the failure of sensor nodes, is presented. A new decentralised Bayesian network based mobility algorithm called BayesMob, which uses local neighbour information, was created to coordinate the self-healing motion. The algorithm enables sensor nodes to predict the probability of an increase in coverage given a move in a particular direction to repair coverage holes. In addition, the thesis outlines the development of a WSN simulator. The latter provides a tool for evaluating the performance of mobile WSNs. All mobility strategies and algorithms discussed herein were simulated using the new WSN simulator.

The characteristic of mobility and the ease of deployment make wireless ad hoc networks suitable for a variety of real life applications that cover a wide range from civilian to military purposes. The lack of a xed infrastructure demands all participating nodes to function as end points of a communication session and also to have routing capabilities. The latter allows data packets to be forwarded to nodes in a multi-hop manner and tackles the routing problem when nodes are joining, leaving or moving around within the network topology unexpectedly. At any time nodes need to be able to provide adaptive, optimal and ecient routing solutions. In order to solve the challenging problem of routing in wireless ad hoc networks, this thesis applies methods from nature and, in particular, from the world of honeybee colonies. A new routing protocol design and its implementation, BeeIP, are proposed and tested. Using honeybee foraging and dancing metaphors, the protocol utilizes special packets to discover paths between sources and destinations. Real honeybees constantly monitor the goodness of their findings based on a number of quality factors such as the distance from the hive, the sweetness of the sugar solution, etc. Then, they effciently distribute the future flights following the most optimal path. Focusing on these key concepts, this work investigates the extent to which a range of low-level network parameters can be used to represent and constantly monitor the goodness of the paths. The design uses a new model to map the honeybee dances and to effiently use multiple paths for future data transmissions. This thesis makes a number of novel contributions. Firstly, an extended mapping of the quality factors from nature to networks and a model to utilize them in order to represent and measure the quality of the paths. Next, the use of statistical prediction by considering prior gathered knowledge to detect any possible improvement or deterioration of path quality over time. Finally, a comprehensive comparison with state-of-the-art protocols in the ns-2 network simulator, the results of which show that BeeIP is able to outperform the others under different conditions and, in particular, in networks of high density, rate of mobility and increased data traffi. Therefore, the proposed design is a viable solution for routing in wireless ad hoc networks.

Energy conservation is considered to be one of the key design challenges within resource constrained wireless sensor networks (WSNs) that leads the researchers to investigate energy efficient protocols with some application specific challenges. Dynamic clustering scheme within the deployed sensor nodes is generally considered as one of the energy conservation techniques. However, unbalanced distribution of cluster heads, highly variable number of sensor nodes in the clusters and high number of sensor nodes involved in event reporting tend to drain out the network energy quickly, resulting in unplanned decrease in network lifetime. Performing power aware signal processing, defining communication methods that can provide progressive accuracy and, optimising processing and communication for signal transmission are the challenging tasks. In this thesis, energy efficient solutions are proposed for collaborative sensing and cooperative communication within resource constrained WSNs. A dynamic and cooperative clustering as well as neighbourhood formation scheme is proposed that is expected to evenly distribute the energy demand from the cluster heads and optimise the number of sensor nodes involved in event reporting. The distributive and dynamic behaviour of the proposed framework provides an energy efficient self-organising solution for WSNs that results in an improved network lifetime. The proposed framework is independent of the nature of the sensing type to support applications that require either time-driven sensing, event-driven sensing or hybrid of both sensing types. A cooperative resource selection and transmission scheme is also proposed to improve the performance of collaborative WSNs in terms of maintaining link reliability. As a part of the proposed cooperative nature of transmission, the transmitreceive antennae selection scheme and lattice reduction algorithm have also been considered. It is assumed that the channel state information is estimated at the receiver and there is a feedback link between the wireless sensing nodes and the fusion centre receiver. For the ease of system design engineer to achieve a predefined capacity or quality of service, a set of analytical frameworks that provide tighter error performance lower bound for zero forcing (ZF), minimum mean square error (MMSE) and maximum likelihood (ML) detection schemes are also presented. The dynamic behaviour has been adopted within the framework with a proposed index derived from the received measure of the channel quality, which has been attained through the feedback channel from the fusion centre. The dynamic property of the proposed framework makes it robust against time-varying behaviour of the propagation environment. Finally, a unified framework of collaborative sensing and communication schemes for cooperative WSNs is proposed to provide energy efficient solutions within resource constrained environments. The proposed unified framework is fully decentralised which reduces the amount of information required to be broadcasted. Such distributive capability accelerates the decision-making process and enhances the energy conservation. Furthermore, it is validated by simulation results that the proposed unified framework provides a trade-off between network lifetime and transmission reliability while maintaining required quality of service.

Process Migration techniques have been around since the early days of Networked Operating Systems. However, process migration has not been widely adopted in software architecture design due to performance cost and unfeasible applications domain at that time. The emergence of new mobile computing technology is now the driving force to investigate new migration models that will support mobile applications and will help to optimise the available scarce resources that are a feature of the mobile computing environment. New alternatives to process migration have been developed and are attracting more researchers. These alternatives include remote execution, process cloning, mobile agents, and mobile code systems. They are simpler to implement and more inexpensive in their overhead cost to the execution environment. This thesis investigates the design and a reference implementation of a framework to support application designers in building adaptive migratory applications. The framework is called Camel. For applications to migrate they need to capture/save their states so that they can be restarted at the new location. However, capturing/saving mechanism of low level execution state imposes great performance cost on applications and are not necessarily required for migration. In Camel we argue that applications can migrate successfully with no need for low level execution state and with minimum application state, if the application designer decides on the necessary state at design time. The design of adaptive migratory applications requires infrastructure support for monitoring the execution environment so that applications can be notified of any changes in a critical resource that might trigger migration, adaptation or a combination of both. Monitoring a critical resource is accomplished through the use of the monitor component of Camel. In this work we only focus on monitoring of memory and application throughput. The framework provides support for the application designer to construct adaptation at design time through the provision of a common framework of helper objects and classes. The designer can then tweak the behaviour of the application through a set of policy profiles to better utilise critical resources

The Radio Frequency Identification (RFID), which includes passive and active systems and is the hottest Auto-ID technology nowadays, and the wireless sensor network (WSN), which is one of the focusing topics on monitoring and control, are two fast-growing technologies that have shown great potential in future logistics management applications. However, an information system for logistics applications is always expected to answer four questions: Who, What, When and Where (4Ws), and neither of the two technologies is able to provide complete information for all of them. WSN aims to provide environment monitoring and control regarded as When and What , while RFID focuses on automatic identification of various objects and provides Who (ID). Most people usually think RFID can provide Where at all the time. But what normal passive RFID does is to tell us where an object was the last time it went through a reader, and normal active RFID only tells whether an object is presenting on site. This could sometimes be insufficient for certain applications that require more accurate location awareness, for which a system with real-time localization (RTLS), which is an extended concept of RFID, will be necessary to answer Where constantly. As WSN and various RFID technologies provide information for different but complementary parts of the 4Ws, a hybrid system that gives a complete answer by combining all of them could be promising in future logistics management applications. Unfortunately, in the last decade those technologies have been emerging and developing independently, with little research been done in how they could be integrated. This thesis aims to develop a framework for the network level architecture design of such hybrid system for on-site resource management applications in logistics centres. The various architectures proposed in this thesis are designed to address different levels of requirements in the hierarchy of needs, from single integration to hybrid system with real-time localization. The contribution of this thesis consists of six parts. Firstly, two new concepts, Reader as a sensor and Tag as a sensor , which lead to RAS and TAS architectures respectively, for single integrations of RFID and WSN in various scenarios with existing systems; Secondly, a integrated ZigBee RFID Sensor Network Architecture for hybrid integration; Thirdly, a connectionless inventory tracking architecture (CITA) and its battery consumption model adding location awareness for inventory tracking in Hybrid ZigBee RFID Sensor Networks; Fourthly, a connectionless stochastic reference beacon architecture (COSBA) adding location awareness for high mobility target tracking in Hybrid ZigBee RFID Sensor Networks; Fifthly, improving connectionless stochastic beacon transmission performance with two proposed beacon transmission models, the Fully Stochastic Reference Beacon (FSRB) model and the Time Slot Based Stochastic Reference Beacon (TSSRB) model; Sixthly, case study of the proposed frameworks in Humanitarian Logistics Centres (HLCs). The research in this thesis is based on ZigBee/IEEE802.15.4, which is currently the most widely used WSN technology. The proposed architectures are demonstrated through hardware implementation and lab tests, as well as mathematic derivation and Matlab simulations for their corresponding performance models. All the tests and simulations of my designs have verified feasibility and features of our designs compared with the traditional systems.

How can an artist-created conversation and data visualisation tool be of value to the practices of data visualisation and social media? In what ways can the processes of designing and analysing such a tool inform the fields of Digital Media Studies, Data Visualisation, Collaboration Studies and Design Methods Research? The primary research was the creation of a unique conversation visualisation tool that encourages creativity in new ways. CodeZebraOS facilitates brainstorming, lateral thought, creative sketching, and the analysis of social dynamics and emotions in online conversation. This thesis integrates dialogues from Digital Media Studies and Collaboration Studies into the field of Data Visualisation. CodeZebraOS contributes a unique metaphor to Data Visualisation, advocating aesthetics that account for relationships between biological, human and virtual worlds, keeping in mind the structure of the source data. CodeZebraOS contributes to current understandings of the complex relationships between human and zoological life forms in Science and Technology disciplines, and applies the study of these relationships to the practice-based design of the tool's aesthetics. A major contribution is the CodeZebra Method, a unique process created for interdisciplinary teams who undertake software development. The methods used to develop CodeZebraOS derive from the empirical approaches of the charrette, participatory design, agile computing and artistic improvisation, creating a hybrid development method. This approach employed usability testing and public deployment at every stage of the tool's development, concluding with a self-reflexive analysis of the artistic and technological results. Finally, this thesis emphasises the skills and theoretical understandings of artists and designers. This thesis argues for the centrality of both the artist's and the designer's role in the creation of social media and data visualisation tools. It makes an original and substantial contribution in four forms: the creation of CodeZebraOS itself; a substantial academic dissertation with an interdisciplinary analysis of practice in the theoretical context of Computer Science (Data Visualisation), Digital Media Studies (ArtSci), Design Methods Research, and Collaboration Studies; the creation and explication of the CodeZebra Method, and finally, a demonstration of the primacy of art and design skills in software development

The CAN (Controller Area Network) protocol provides one of the cost-effective methods to network current generations of distributed embedded systems. Although it is a robust protocol with short messages and simple priorities, it is largely thought of as only being suitable for soft real-time, event-triggered systems. Safety critical applications require highly predictable behaviour with strict bounds on worst-case message transmission times; the next-generation mechatronic systems also requires a high level of information throughput. In its current form, CAN lacks most of these requirements principally due to its medium access scheme and physical-layer design. This thesis presents a frame work which aims to enhance the capabilities of CAN, in order to push the boundaries of the protocol’s current operation. In particular, the main research question to be addressed is the exploration of the extent to which low-level modifications can enhance CAN suitability for use in the next generation of critical systems. In order to answer this question, it is first necessary to develop a flexible and robust platform to implement these modifications using a novel facility made up from custom soft-core CAN controllers. This novel facility was then employed to implement and experimentally investigate three small but conceptually significant protocol modifications as follows: Increasing the effective data rate from 1 to 10 Mbps whilst doubling the effective payload from 8 to 16 bytes; Reduction of unwanted transmission jitter by compensating for bit stuffing; Enabling a windowed transmission scheme to provide optimal trade-off’s between transmission reliability and real-time behaviour in noisy environments. The thesis describes the results obtained from these experiments and summarizes the main pros and cons that appear. The thesis then concludes with observation that the modified CAN protocol may be suitable for use with certain classes, of the next generation time-critical distributed embedded systems.

Parallel computers are generally either shared-memory machines or distributed- memory machines. There are currently technological limitations on shared-memory architectures and so parallel computers utilizing a large number of processors tend tube distributed-memory machines. We are concerned solely with distributed-memory multiprocessors. In such machines, the dominant factor inhibiting faster global computations is inter-processor communication. Communication is dependent upon the topology of the interconnection network, the routing mechanism, the flow control policy, and the method of switching. We are concerned with issues relating to the topology of the interconnection network. The choice of how we connect processors in a distributed-memory multiprocessor is a fundamental design decision. There are numerous, often conflicting, considerations to bear in mind. However, there does not exist an interconnection network that is optimal on all counts and trade-offs have to be made. A multitude of interconnection networks have been proposed with each of these networks having some good (topological) properties and some not so good. Existing noteworthy networks include trees, fat-trees, meshes, cube-connected cycles, butterflies, Möbius cubes, hypercubes, augmented cubes, k-ary n-cubes, twisted cubes, n-star graphs, (n, k)-star graphs, alternating group graphs, de Bruijn networks, and bubble-sort graphs, to name but a few. We will mainly focus on k-ary n-cubes and (n, k)-star graphs in this thesis. Meanwhile, we propose a new interconnection network called augmented k-ary n- cubes. The following results are given in the thesis.1. Let k ≥ 4 be even and let n ≥ 2. Consider a faulty k-ary n-cube Q(^k_n) in which the number of node faults f(_n) and the number of link faults f(_e) are such that f(_n) + f(_e) ≤ 2n - 2. We prove that given any two healthy nodes s and e of Q(^k_n), there is a path from s to e of length at least k(^n) - 2f(_n) - 1 (resp. k(^n) - 2f(_n) - 2) if the nodes s and e have different (resp. the same) parities (the parity of a node Q(^k_n) in is the sum modulo 2 of the elements in the n-tuple over 0, 1, ∙∙∙ , k - 1 representing the node). Our result is optimal in the sense that there are pairs of nodes and fault configurations for which these bounds cannot be improved, and it answers questions recently posed by Yang, Tan and Hsu, and by Fu. Furthermore, we extend known results, obtained by Kim and Park, for the case when n = 2.2. We give precise solutions to problems posed by Wang, An, Pan, Wang and Qu and by Hsieh, Lin and Huang. In particular, we show that Q(^k_n) is bi-panconnected and edge-bipancyclic, when k ≥ 3 and n ≥ 2, and we also show that when k is odd, Q(^k_n) is m-panconnected, for m = (^n(k - 1) + 2k - 6’ / ‘_2), and (k -1) pancyclic (these bounds are optimal). We introduce a path-shortening technique, called progressive shortening, and strengthen existing results, showing that when paths are formed using progressive shortening then these paths can be efficiently constructed and used to solve a problem relating to the distributed simulation of linear arrays and cycles in a parallel machine whose interconnection network is Q(^k_n) even in the presence of a faulty processor.3. We define an interconnection network AQ(^k_n) which we call the augmented k-ary n-cube by extending a k-ary n-cube in a manner analogous to the existing extension of an n-dimensional hypercube to an n-dimensional augmented cube. We prove that the augmented k-ary n-cube Q(^k_n) has a number of attractive properties (in the context of parallel computing). For example, we show that the augmented k-ary n-cube Q(^k_n) - is a Cayley graph (and so is vertex-symmetric); has connectivity 4n - 2, and is such that we can build a set of 4n - 2 mutually disjoint paths joining any two distinct vertices so that the path of maximal length has length at most max{{n- l)k- (n-2), k + 7}; has diameter [(^k) / (_3)] + [(^k - 1) /( _3)], when n = 2; and has diameter at most (^k) / (_4) (n+ 1), for n ≥ 3 and k even, and at most [(^k)/ (_4) (n + 1) + (^n) / (_4), for n ^, for n ≥ 3 and k odd.4. We present an algorithm which given a source node and a set of n - 1 target nodes in the (n, k)-star graph S(_n,k) where all nodes are distinct, builds a collection of n - 1 node-disjoint paths, one from each target node to the source. The collection of paths output from the algorithm is such that each path has length at most 6k - 7, and the algorithm has time complexity O(k(^3)n(^4)).