In this thesis we propose the use of network coding to improve the energy efficiency in core networks, by reducing the resources required to process traffic flows at intermediate nodes. We study the energy efficiency of the proposed scheme through three approaches: (i) developing a mixed integer linear programme (MILP) to optimise the use of network resources. (ii) developing a heuristic based on minimum hop routing. (iii) deriving an analytical bounds and closed form expressions. The results of the MILP model show that implementing network coding over typical networks can introduce savings up to 33% compared to the conventional architectures. The results of the heuristic show that the energy efficient minimum hop routing in network coding enabled networks achieves power savings approaching those of the MILP model. The analytically calculated power savings also confirm the savings achieved by the model. Furthermore, we study the impact of network topology on the savings obtained by implementing network coding. The results show that the savings increase as the hop count of the network topology increases. Using the derived expressions, we calculated the maximum power savings for regular topologies as the number of nodes grows. The power savings asymptotically approach 45% and 23% for the ring (and line) and star topology, respectively. We also investigate the use of network coding in 1+1 survivable IP over WDM networks. We study the energy efficiency of this scheme through MILP, a heuristic with five operating options, and analytical bounds. We evaluate the MILP and the heuristics on typical and regular network topologies. Implementing network coding can produce savings up to 37% on the ring topology and 23% considering typical topologies. We also study the impact of varying the demand volumes on the network coding performance. We also develop analytical bounds for the conventional 1+1 protection and the 1+1 with network coding to verify the results of the MILP and the heuristics and study the impact of topology, focusing on the full mesh and ring topologies, providing a detailed analysis considering the impact of the network size.

Concurrent and distributed behaviour encompasses a wide range of ever evolving phenomena and features of computation such as communication, mobility, causality, failure recovery and reversibility. In order to understand better and make precise the properties of such behaviour, concurrent and distributed behaviour needs to be modelled abstractly and with formal rigour. Delay-insensitive networks are a class of asynchronous systems which makes no assumptions about the timing of signals or components. This makes them suitable for the implementation of highly concurrent systems. Unfortunately, concurrency within delay-insensitive networks is an underdeveloped concept which lacks formal rigour. Reversibility of such systems is also typically only studied in the context of serial systems without concurrency. In this thesis, a new model for describing the behaviour of delay-insensitive components is introduced which more naturally permits the modelling and study of concurrency. Reversibility of such components is discussed. The concept of an environment for a component is formalised, and its limitations in terms of interactivity with such a component is also studied. Algorithms for generating the environments for delay-insensitive components, such that desirable properties always hold are given. Universality results and properties of networks of such components are examined in-depth. A new process algebra which allows the encoding of these networks is introduced. This permits rigorously defined notions of implementation and other desirable run-time properties of these networks. A family of new novel cellular automata is defined which allows the encoding of delay-insensitive networks. These cellular automata are competitive with existing CA regarding universality, and number of rules and states. They have a feature we call direction-reversibility, which allows the inversion of behaviour simply by reversing the direction of signals. Finally, two pieces of software called Delay-Insensitive Network Tool Suite and STCA Simulator, developed to aid in this research, are also detailed.

Mobile Ad-Hoc Networks (MANETs) are known for their versatility, which is they are capable of supporting many applications. In addition to this versatility MANETs are quick to deploy without need for an existing predefined communications infrastructure. However, although the lack of infrastructure allows for the quick deployment of the data communications network, it adds many factors that hinder packet delivery. Such hindrances occur because of the dynamic topology caused by the mobility of the nodes which results in link breakages. Routing protocols exist that attempt to refresh available routes; however, this is after link breakages have occurred. The nodes also usually have constrained resources (i.e. energy source and limited bandwidth). This thesis presents a novel approach of network behaviour and management by implementing cognitive attributes into a MANET environment. This allows an application to better meet its mission objectives, decreases the end-to-end delay, and increases packet delivery ratio. The network is able to make observations, consider previous actions and consequences of the actions, and make changes based on the prior knowledge and experience. This work also shows how the network can better utilise limited resources such as bandwidth allocation by applying cognitive attributes. Simulations conducted show promising results and prove that an increase in network performance is possible if adopting a cross-layered approach and allow the network to manage and to ‘think’ for itself. Various simulations were run with various scenarios and results are presented without cognition applied, with partial cognition applied and with full cognition applied. A total of 52 simulations were run and from this the results were compared and contrasted. The analysis shows that cognitive attributes does increase network performance in the majority of applications.

The ubiquity of multihoming amongst mobile devices presents a unique opportunity for users to co-operate, sharing their available Internet connectivity, forming multihomed mobile networks on demand. This model provides users with vast potential to increase the quality of service they receive. Despite this, such mobile networks are typically underutilized and overly restrictive, as additional Internet connectivity options are predominantly ignored and selected gateways are both immutable and incapable of meeting the demand of the mobile network. This presents a number of research challenges, as users look to maximize their quality of experience, while balancing both the financial cost and power consumption associated with utilizing a diverse set of heterogeneous Internet connectivity options. In this thesis we present a novel architecture for mobile networks, the contribution of which is threefold. Firstly, we ensure the available Internet connectivity is appropriately advertised, building a routing overlay which allows mobile devices to access any available network resource. Secondly, we leverage the benefits of multipath communications, providing the mobile device with increased throughput, additional resilience and seamless mobility. Finally, we provide a multihomed framework, enabling policy driven network resource management and path selection on a per application basis. Policy driven resource management provides a rich and descriptive approach, allowing the context of the network and the device to be taken into account when making routing decisions at the edge of the Internet. The aim of this framework, is to provide an efficient and flexible approach to the allocation of applications to the optimal network resource, no matter where it resides in a mobile network. Furthermore, we investigate the benefits of path selection, facilitating the policy framework to choose the optimal network resource for specific applications. Through our evaluation, we prove that our approach to advertising Internet connectivity in a mobile network is both efficient and capable of increasing the utilization of the available network capacity. We then demonstrate that our policy driven approach to resource management and path selection can further improve the user’s quality of experience, by tailoring network resource usage to meet their specific needs.

Since its inception, the World Wide Web has revolutionised the way we share information, keep in touch with each other and consume content. In the latter case, it is now used by thousands of simultaneous users to consume video, surpassing physical media as the primary means of distribution. With the rise of on-demand services and more recently, high-definition media, this popularity has not waned. To support this consumption, the underlying infrastructure has been forced to evolve at a rapid pace. This includes the technology and mechanisms to facilitate the transmission of video, which are now offered at varying levels of quality and resolution. Content delivery networks are often deployed in order to scale the distribution provision. These vary in nature and design; from third-party providers running entirely as a service to others, to in-house solutions owned by the content service providers themselves. However, recent innovations in networking and virtualisation, namely Software Defined Networking and Network Function Virtualisation, have paved the way for new content delivery infrastructure designs. In this thesis, we discuss the motivation behind OpenCache, a next-generation content delivery platform. We examine how we can leverage these emerging technologies to provide a more flexible and scalable solution to content delivery. This includes analysing the feasibility of novel redirection techniques, and how these compare to existing means. We also investigate the creation of a unified interface from which a platform can be precisely controlled, allowing new applications to be created that operate in harmony with the infrastructure provision. Developments in distributed virtualisation platforms also enables functionality to be spread throughout a network, influencing the design of OpenCache. Through a prototype implementation, we evaluate each of these facets in a number of different scenarios, made possible through deployment on large-scale testbeds.

No description available.

This thesis is principally concerned with effcient energy consumption in wireless sensor networks from two distinct aspects from a theoretical point of view. The thesis addresses the issue of reducing idle listening states in a restricted tree topology to minimise energy consumption by proposing an optimisation technique: the extra-bit technique. This thesis also focuses on showing lower bounds on the optimal schedule length, which are derived for some special cases of the tree, such as a single chain, balanced chains, imbalanced chains, three and four level k-ary trees and Rhizome trees. Then, we propose an algorithm which can exactly match the lower bound for a single chain, balanced chains and Rhizome trees individually and which is a few steps away from the optimal solution for imbalanced chains. Finally, we propose the use of two frequencies to further save energy and minimize latency. Recent research has shown that significant energy improvements can be achieved in WSNs by exploiting a mobile sink for data collection via single hop communications. A mobile sink approaches the transmission range of sensors to receive their data and deposit the data at the base station. The thesis, as a second problem, focuses on the design issues of an energy efficient restricted tour construction for sink mobility. We propose two different techniques. The first one is heuristic and uses a criterion based on maximum coverage and minimum energy consumption called the "max-ratio". Although its time complexity is polynomial, this heuristic algorithm cannot always produce a good solution. As a result, we propose the sec- ond algorithm. Despite the time complexity of the second algorithm being pseudo polynomial, the optimal solution can be found if one exists. For each algorithm men- tioned, two scenarios are taken into account with regard to the transmission. In the first scenario, one assumes that there is no upper bound on the transmission range while in the second setting the nodes can adjust their transmission range between 0 and the maximum range. The algorithms have been implemented and simulated in Matlab.

Networks are everywhere, changing the way we communicate with each other, transport goods and share information. The problems of efficient operation of such networks can often be stated as (abstract) network flow problems. In a problem of this type we want to send some commodity (goods, messages, data, electricity, vehicles) from supply points to demand points in an underlying network, which is modeled as a graph. There are various constraints on the characteristics of the routes, such as capacities and costs. There may be a number of different optimization objectives, depending on the problem setting. Network flow problems form one of the most important and most frequently encountered classes of optimization problems. They lie at the intersection of several scientific fields including computer science, mathematics and operational research. We are interested in the computer science aspect of network optimization problems, that is, in development and analysis of efficient algorithms for such problems. In this thesis we study algorithmic frameworks for multicommodity flow problems, which can be described in the following way. The input is a directed graph G = (N; E), where N is the set of nodes and E is the set of edges, and specifications of k commodities. Each edge has an associated capacity c(e) and each commodity has an associated source-sink pair of nodes (si; ti) and a demand value di. The goal is to design simultaneous flow of all commodities that satisfies their demands, takes into account the capacities of the edges and optimizes a specified objective function. We focus on the problem of minimizing the overall congestion, which is often referred to as the Maximum Concurrent Flow problem. We consider both sequential and distributed models of computation. We show that the two main sequential algorithmic Maximum Concurrent Flow frameworks - the rerouting framework and the incremental framework - are more closely related than previously assumed. We prove that the running time of some distributed Maximum Concurrent Flow algorithms shown recently are asymptotically tight. We also propose a heuristic for these algorithms to improve their performance on some types of inputs.

A Mobile Ad-Hoc Network (MANET) is a collection of randomly distributed infrastructure-less mobile nodes that form a wireless network. These Mobile nodes have the capability to act as a host or relay. As a host, the mobile nodes can be the source and/or destination of traffic, and when acting as a relay, they can be an intermediate node that forwards the traffic to its destination. Some of the challenges of a MANET include the dynamic network topology, device discovery, power constraints, wireless channel conditions and limited network resources. These challenges degrade the network performance and thus affect the network stability and robustness. Therefore, it is difficult for a MANET to attain the Quality of Service (QoS) of a wired network. This thesis aims to address the problem of the limited wireless network resources by proposing two adaptive scheduling algorithms that can adapt in real-time to the changes in the network. To achieve the aim; this thesis first analyses the behaviour of various application profiles in a queue. It models Voice, Email, and Internet Browsing traffic (by specifying packet sizes, and inter-arrival rates based on various distributions) separately and then simultaneously in a common network for uncongested and congested conditions, after which scheduling is applied in order to improve the overall network performance. The Voice traffic profile is then added to the UDP/IP protocol stack and the network performance is compared to a simple node without the UDP/IP protocol stack. A realistic wireless propagation model for the simulation is developed from a point-to-point open-field outdoor experiment. This thesis proposes two adaptive priority fuzzy based scheduler for a MANET, the priority of packets in the queue are determined based on the real-time available network resources. The methodology for transmitting a live-feed video stream over OPNET to validate the scheduler is also presented. An interface between the simulation and hardware is created to send real-time video traffic through the simulation network. This thesis concludes by showing that the performance of a MANET network can be improved by applying an adaptive scheduler.

This thesis is a description of the research undertaken in the course of the PhD and evolves around a context transfer protocol which aims to complement and support mobility management in next generation mobile networks. Based on the literature review, it was identified that there is more to mobility management than handover management and the successful change of routing paths. Supportive mechanisms like fast handover, candidate access router discovery and context transfer can significantly contribute towards achieving seamless handover which is especially important in the case of real time services. The work focused on context transfer motivated by the fact that it could offer great benefits to session re-establishment during the handover operation of a mobile user and preliminary testbed observations illustrated the need for achieving this. Context transfer aims to minimize the impact of certain transport, routing, security-related services on the handover performance. When a mobile node (MN) moves to a new subnet it needs to continue such services that have already been established at the previous subnet. Examples of such services include AAA profile, IPsec state, header compression, QoS policy etc. Re-establishing these services at the new subnet will require a considerable amount of time for the protocol exchanges and as a result time- sensitive real-time traffic will suffer during this time. By transferring state to the new domain candidate services will be quickly re-established. This would also contribute to the seamless operation of application streams and could reduce susceptibility to errors. Furthermore, re-initiation to and from the mobile node will be avoided hence wireless bandwidth efficiency will be conserved. In this research an extension to mobility protocols was proposed for supporting state forwarding capabilities. The idea of forwarding states was also explored for remotely reconfiguring middleboxes to avoid any interruption of a mobile users' sessions or services. Finally a context transfer module was proposed to facilitate the integration of such a mechanism in next generation architectures. The proposals were evaluated analytically, via simulations or via testbed implementation depending on the scenario investigated. The results demonstrated that the proposed solutions can minimize the impact of security services like authentication, authorization and firewalls on a mobile user's multimedia sessions and thus improving the overall handover performance.