In this thesis, I visualised landscapes of urban Wi-Fi (wireless connection) networks as a hybrid of art and anthropology. The process involved 'how the subjects are defined in Wi-Fi networks,' 'how the data is collected in cities' and 'how the conversion between Wi-Fi networks and artistic landscapes takes place.' The subject of my project is Wi-Fi access points, which allow users to connect to and access the internet. I viewed Wi-Fi users as cyborgs (cybernetic organisms) because they entered cyberspace with their Wi-Fi-equipped computers or phones. Access points and the users' devices as myths are necessary elements used in immersing themselves in different worlds. Claude Levi-Strauss considered myths as bricolage in organising heterogeneous events to create structures; that is, people can explain the occurrence of events in daily life with structures. Wi-Fi users' wireless equipment represents an internet-enabled structure to help users go online. In other words, Wi-Fi users' identities are tied to their particular structures - Wi-Fi equipment and the devices used to access it. In addition, the mixture of flesh and machine contributes to defining human users as cyborgs. I transformed the access points of the cyborgs' houses using metaphor and the house concept inspired by Levi-Strauss's house society theory. Houses, as places where people live, form basic economic/political units. Both invisible titles and visible, physical buildings forge the identities of the houses' occupants. In my works, access points have similar features, so I used houses as the source metaphor. With common attributes and functions, the house metaphor creates a connection between different objects (houses and access points) on the same level. All Wi-Fi access points have a unique 12-character identifier code composed of numbers (from 0 to 9) and letters (from a to f). As I walked through different cities along specific collection routes as a native Wi-Fi cyborg, I collected the Wi-Fi access points' identifier codes. Time was a unique element that added to my works, in that the sequence of time-related variables (including signal strength, new Wi-Fi APs, and abandoned APs) could be experienced only once. Following data collection, the identifier codes were converted to colour grids and arranged along the collection route to depict Wi-Fi landscapes. Although the mapping was carried out by computers, the collection routes formed the personal features of landscapes, as in Impressionist paintings. The use of colour grids/squares was widespread in modernist artworks, such as those by Josef Albers and Donald Judd. These artists pursued a notion of pure art by creating isolated colour compositions with simple geometrical shapes in their works. As relates to Lucy Lippard's terms 'dematerialization' and 'deemphasis on material aspects,' some postmodemist artists, such as Liam Gillick, applied modernist art language in specific contexts, rather than isolating their works as closed systems. Art for postmodemists was the process and context of the creation as opposed to the objects represented. The colour grids in my work are interwoven creations that juxtapose the hybridity of modernism and postmodemism. The colours relate to social contexts because they come from access points along the streets of different cities; moreover, the colours interact with each other to create a relationship first proposed by Albers. My works aim to reflect that our current situation, as it relates to technology, is juxtaposed by different hybrids, including modemism/postmodernism and humanismlposthumanism. Wi-Fi landscapes in this thesis did Dot contain all access points because Wi-Fi devices are still growing in cities. In other words, one can only present the complete features of contemporary societies via their incomplete and heterogeneous composition.

In recent times, broad band Internet connectivity has become something presumed accessible to all, often shared throughout the home and between multiple users and devices. Despite the proliferation of online services this perceived ubiquity is unfortunately false, many are still unable to receive high quality Internet access within their homes, due to infrastructure restrictions or geographical problems. One solution is through the deployment of community networks to share Internet access, initiated, designed, and managed by ordinary people with little or no technical background. This thesis takes an interdisciplinary approach to understanding the challenges faced by the users of such networks, and emphasises the importance of user focussed techniques when designing network management solutions for community settings. It investigates the ways in which communal networks can be improved by encouraging more formalised resource sharing, and how users can be aided to better understand their network usage through the design and implementation of an appropriate system.

The future success of high frequency (HF) communication systems rely on its ability to integrate and support IP diversity within a multiple intemet protocol (IP) based networks, such as satellite communication (SATCOM), local area network (LAN), wide area network (WAN) bearers. The introduction of new and proposed standards on HP-IP in recent years has increased the interest in the areas of performance analysis of HP -IP communication systems and networks. A wide range of modem services rely on IP and current HP-IP systems can support 2.4 to 19.2 kbps services such as e-mail and intemet. However, the reliability and the quality of service (QoS) still remains an issue of interest, particularly over longer distance skywave channels. These modem services require a higher data rate, much better bandwidth utilisation and a good QoS for its successful implementation. This work investigated HP-IP systems with the aim of improving the performance of legacy, current and proposed future systems without modifications to existing hardware systems. Initially the research conducted involved practical measurements and analysis on HF-IP systems complying with proposed NATO STANAG 5066 draft/edition 2 standards. Having investigated several NATO HF-IP standards (STANAG 5066 editionl, STANAG 5066 draft/edition 2, STANAG 4539/4285/4529, etc), a novel concept of error control coding (ECC) within the data link (DL) layer for HP-IP systems was proposed. Benefit of this proposed concept is that it does not require hardware modifications in legacy and current system for improving the performance. For application of this concept high performance low density parity check (LDPC) coding was considered. Two classes of short block length quasi-cyclic (QC) LDPC codes with switchable- rate single encoder/decoder structure; based on finite fields were designed and constructed. Several code rates were constructed within a single encoder/decoder structure resulting in reduced implementation complexity. Both classes of codes were simulated using HF channel model (ITU- R F.1487) covering latitudes and conditions for performance analysis. The simulation results show by using switchable-rate QC-LDPC coding scheme that there is coding gain of 2.4 dB compared to the existing STANAG 4539 convolutional coding scheme demonstrating the high performance of the proposed scheme in ITU-R F.l487 HF channel environment. In addition, the use of STANAG 5066 draft/edition 2 operating on a skywave multi-node HF-IP token ring (TR) WAN for a civilian disaster relief scenario was investigated. Here, a novel HF-IP network concept was proposed. The concept incorporates multi-node HF-IP TR WAN as an inner network, supported by an outer network made up of digital radio monodiale (DRM) service operating on a single frequency within the HF band. As STANAG 5066 draft/edition 2 was primarily designed to supporting multi-node HF-IP networks, it was vital to understand the network reliability and number of practical nodes that this network can support in different skywave HF channel conditions. A 3-node network based on skywave propagation covering a large geographical area was investigated. Using this scenario probability of reliability of a skywave multi-node HF-IP was analysed by simulations and practical measurements using STANAG 5066 draft/edition 2 IP protocol and STANAG 4539 modem setups. This analysis showed that the skywave multi-node HF-IP TR network can reliably operate between 3-5 nodes.

Advances in the mobile device manufacture, rapid growth of Web services development and progression of wireless communication with the widespread use of Internet applications are the most recent trends in distributed information systems. The evolution of these trends yields the emergence of Mobile Web Services technology (MWS). Adaptive service provision and execution in mobile environments is a new avenue in MWS. It is emanated from the need to cope with the existence of mobile resource limitations to allow the reliable provision of a number of useful services that are hosted on these mobile devices. This research argues that the mechanisms used to facilitate service distribution allow non-interrupted provision of MWS efficiently. The objective of this research is to investigate these mechanisms and define a system for applying them. The main criteria for this system are flexibility, dynamicity and transparency. Simple Object Access Protocol (SOAP)-based Mobile Host Web service Framework (MHWF) is reproduced and extended as part of this research, to allow deploying, providing and executing SOAP-based services. Correspondingly, Representational State Transfer (RESTful) MHWF is defined and developed for providing RESTful-based services. Both frameworks have been analysed and compared in terms of performance, scalability, reliability and resource consumptions. Moreover, they both have been extended to allow service distribution through offloading, which is the first explored distribution mechanism. RESTful-based technology has shown that it is more convenient for mobile environment. This research also classifies the distribution strategies into three classes: Contentment Distribution (CD), Simple Partial Distribution (SPD) and Complex Partial Distribution (CPD). The distinction emerges from variance in the types and complexity levels of services that influence the quantity and quality of the distribution mechanisms' usage. Novel approaches are proposed in order to exploit these mechanisms and to define and setup the building blocks for the corresponding MHWFs. The correct behaviour of these frameworks is empirically validated and their safety properties are also verified analytically using formal methods. This is complemented by a proof of concept demonstration. Furthermore, an evaluation of their performance is carried out by simulation. These evaluation results are interpreted as Fuzzy logic rules that are used to trigger and control distribution schemes. Last but not least, an innovative approach to partition and orchestrate the execution tasks of the distributed services is followed based on the hierarchical structure used for representing the Uniform Resource Identifier (URI) of the invoked services

The trend of moving conventionallP networks towards Next Generation Networks (NGNs) has highlighted the need for more sophisticated Traffic Managers (TMs) to guarantee better network and Quality of Service (QoS); these have to be scalable to support increasing link bandwidth and to cater for more diverse emerging applications. Current TM solutions though, are limited and not flexible enough to support new TM functionality or QoS with increasing diversity at faster speeds. This thesis investigates efficient and flexible memory management architectures that are critical in determining scalability and upper limits of TM performance. The approach presented takes advantage of current FPGA technology that now offers a high density of computational resources and flexible memory configurations, leading to what the author contends to be an ideal, programmable platform for distributed network management. The thesis begins with a survey of current TM solutions and their underlying technologies/architectures, the outcome of which indicates that memory and memory interfacing are the major factors in determining the scalability and upper limits of TM performance. An analysis of the implementation cost for a new TM with the capability of integrated queuing and scheduling further highlights the need to develop a more effective memory management architecture. A new on-demand QM architecture for programmable TM is then proposed that can dynamically map the ongoing active flows to a limited number of physical queues. Compared to the traditional QMs, it consumes much less memory resources, leading to a more scalable and effiCient TM solution. Based on the analysis of the effect of varying Internet traffic on the proposed OM, a more robust and resilient QM architecture is derived that achieves higher scalability and pefformance by adapting its functionality to the changing network conditions.

Support for IP-based services, efficient resource management and broadband  connectivity distinguish next generation of satellite systems from their narrowband  predecessors. Two types of satellite access systems are studied: Hybrid and Bandwidth-on-Demand (BoD). The capacity asymmetry in these types of networks interacts with  TCP resulting in reduction of forward path performance due to the characteristics of die return path. This thesis presents an ACK Modification scheme combining ACK Filtering with uplink Fair Queuing to improve the fair sharing and forward diroughput in the presence of bi-directional and competing traffic. DVB-RCS is a standard that describes an example of two-way BoD satellite access network technology. BoD is a  dynamic resource allocation mechanism to optimise resource management. While increasing transmission efficiency, BoD schemes may introduce large delay, delay variations and capacity variations that impact die performance of TCP. This thesis identifies BoD characteristics that impacts TCP and present a set of rules for designing BoD to support TCP. An implementation of different QoS classes using BoD techniques to provide loose delay guarantees for TCP traffic is also studied. This thesis compares a range of enhancement techniques described in the literature to overcome the effect of large path delay and show mat the requirement for split TCP solutions in satellite networks could be eliminated by combining the connection caching techniques adopted by the Congestion Manager and HTTP1.1 with Applications layer middle boxes (e.g. web proxies). A GEO BoD system may also impact the protocol behaviour. The Current TCP timer algorithm fails to operate correctly when initial network path delay exceeds 3 seconds. This thesis presents an alternative mechanism to choose initial RTO value for TCP data transfer and define the congestion control behaviour during TCP start-up.

No description available.

No description available.

No description available.

Web Services form the basis of the web based eCommerce eScience applications so it is vital that robust services are developed. Traditional validation and verification techniques are centred around the concept of removing all faults to guarantee correct operation whereas Dependability gives an assessment of how dependably a system can deliver the required functionality by assessing attributes, and by eliminating threats via means attempts to improve dependability. Fault injection is a well-proven dependability assessment method. Although much work has been done in the area of fault injection and distributed systems in general, there appears to have been little research carried out on applying this to middleware systems and Web Services in particular. There are additional problems associated with applying existing fault injection technologies to Web Services running in a virtual machine environment since most are either invasive or work at a machine level. The Fault Injection Technology (FIT) method has been devised to address these problems for middleware systems. The Web Service-Fault Injection Technology (WS-FIT) implementation applies the FIT method, based on network level fault injection, to Web Services to create a non-invasive dependability assessment method. It allows targeted perturbation of Web Service RFC parameters as well as more traditional network level fault injection operations. The WS-FIT tool includes taxonomies that define a system under test, fault models to apply and failure modes to be detected, and uses these taxonomies to generate fault injection campaigns. WS-FIT has been applied to a number of case studies and has successfully demonstrated its effectiveness. It has also been successfully applied to a third-party system to evaluate dependability means. It performed this dependability assessment as well as allowing debugging of the means to be undertaken uncovering unknown faults.