Thesis (Ph. D.)--University of Texas at Austin, 2007. / Vita. Includes bibliographical references.

Thesis (Ph. D.) -- University of Texas at Dallas, 2007. / Includes vita. Includes bibliographical references (leaves 78-80)

Thesis (Ph. D.)--Ohio State University, 2007. / Title from first page of PDF file. Includes bibliographical references (p. 142-147).

The frequency assignment problem involves the assignment of discrete channels (frequencies) to the transmitters of a radio network. A separation between the frequencies assigned to transmitters close to each other is required to avoid interference. Unnecessary separation causes an excess requirement for spectrum, which is a valuable resource. Consequently good assignments minimise both interference and the spectrum required. The subject of this thesis is the fixed spectrum frequency assignment problem, where the spectrum available is given and the target is to minimize the total interference of the system. Interference is modelled through binary constraints, and consequently the problem, which is treated as a combinatorial optimisation problem, can be represented by an undirected weighted graph. A summary of some of the integer programming formulations which model the problem is presented, together with a brief dimensional study of them. An efficient implementation of two well-known metaheuristic algorithms, adapted to the problem treated, is described. Some novel lower bounding techniques which, given a problem, work by combining lower bounds calculated for some of its clique-like subproblems are presented. The key idea is that it is quite easy to calculate tight lower bounds for problems represented by complete graphs (cliques). The lower bounds for clique-like subproblems are produced by two different methods, the first of which is based on the solution of a linear program, while the second is based on a closed formula. The most effective method to generate estimates for general problems is based on a linear program which is reinforced with inequalities derived from the lower bounds calculated on its clique-like subproblems. The last part of the thesis is dedicated to improvements to the lower bounding techniques, both for those working on general problems and for those developed for cliques only. Detailed computational results, obtained on a wide range of benchmarks, are reported.

Hydrogen technology has widely been acknowledged as a route to reduced carbon emissions and energy security. However, it is also recognised that technology development alone is not enough to cause a shift change in uptake and that public perception also needs to be considered. This paper identifies the public perception of hydrogen technology (with specific interest in biologically produced hydrogen) and determines whether it meets the public expectation in terms of its potential energetic output. The energetic viability of hydrogen (defined for this purpose as a net positive energy balance) is derived by determining the potential for agriculturally produced hydrogen using dark fermentative production techniques. A suitable range of energy crops have been identified using selection criteria including yield, harvest window and composition of the crops. Public perception of hydrogen energy was investigated through the use of two focus groups. This identified Welsh public opinions on the production and end use of hydrogen energy. Due to the rapid progress in wireless communication technologies and the increasing demands for new services; cutting edge techniques and research have developed wireless access delivery of broadband data. Such systems referred to as Broadband Fixed Wireless Access Systems (BFWA) or alternatively, Local Multipoint Distribution Systems (LMDS), are increasingly being regarded as a legitimate challenger to cable and digital subscriber lines (DSL), particularly in markets with poor cable or copper infrastructure. These systems have capabilities that go beyond the current WiFi and WiMAX technologies by providing users with higher bandwidths and faster data rates. The frequency of operation for such networks lies between 28 and 42 GHz according to the spectrum bands allocated for future LMDS networks. The presented work aims at proposing a generalized stochastic model for the LMDS urban/suburban propagation channel. The model is based on a physical electromagnetic representation of the millimetre wavelength channel, with particular emphasis on fading margins in line-of-sight (LOS) links. The study focuses on the analysis, modelling and measurements of the fading effects of signal scattering caused by building surfaces and the ground on the direct field in an LMDS link at millimetre­ waves. Outcomes have been validated against experimental results obtained in realistic propagation scenarios. More than 180 field measurement sets were taken at 40 GHz for model testing and validation purposes. The main innovation of this work is the solution proposed to address the problem of signal scatter at millimetre-wave lengths. The approach uses high-frequency approximations to the analytical solution given by Kirchhoff s tangent-plane representation of rough surfaces. The electromagnetic field and signal power at the receiver are considered random and are evaluated using the Physical Optics method over different possible realizations of the surface geometry and building architectural features. This has resulted in novel derivations of the mean field, mean power density and scatter distribution of the scattered field. In comparison to deterministic models, such as ray-tracing that require intensive computations and detailed (millimetric resolution) knowledge of topographical data that are almost impossible to obtain, the proposed model is rather efficient and yields accurate results. The model has also been employed in analysing the effects of building scatter on the variance of the main desired link and the adjacent- or co-channel interfering link arising from neighbouring cells in cellular LMDS networks. This has enabled predictions of signal-to-interference ratio statistics and distributions for a particular underlying propagation environment. Overall, compansons yielded very good agreements between measurements and predictions of the main statistical parameters, thus verifying the main assumptions relating to the received signal strength, as well as the validity of a Rician distribution in describing the signal envelope variability for both the main and interference links in complex millimetre-wave propagation environments.

Modern telecommunication systems and standards are mainly dependent on the availability of digital signal processing capabilities of appropriate hardware components. Two main categories can be distinguished in the development of digital signal processing units. On the one hand a number of general purpose digital signal processors are available on the market, which can be programmed through programming languages like C or C++ or - for higher performance purpose - directly in assembly code. The advantage of such devices is the ability of high flexibility and short time to market since there is no further hardware development, on the integrated circuit level, required. On the other hand hardware components are specifically developed for signal processing tasks, which are mainly application specific integrated circuits (ASIC). They are usually only programmable to a certain degree, always considering the area of application, i.e. wireless telecommunication systems. Although they do not offer the flexibility of general purpose digital signal processors, they offer the big advantage of less required hardware (measured as chip area or die size), lower power consumption and higher speeds. Usually hybrids are found on the market, which combine freely programmable Digital Signal Processor (DSP) with very specific hardware modules to support the specific application needs. This thesis describes the development of a Wireless Telecommunication System, describing the relevant development methodologies, regarding aspects of hardware and software split and actual implementations of components in hardware as well as in software. This is done specifically for the example of a wideband code division multiple access (WCDMA) wireless mobile system. The actual state of the art is described in detail, according to the relevant literature in the area of WCDMA systems. Programmable hardware is presented, which is covered through a portfolio of patents. The purpose and the application of each patent are described in detail as well as the area of application. Finally a classification of each patent is given, which aims to give an objective measure about the value of a patent. The presented patents show a significant contribution to knowledge enabling the development of low power mobile wireless telecommunication systems.

D.Ing. / This thesis sets out to propose and analyze a cellular Direct Sequence Spread Spectrum Multiple Access (DSjSSMA) system for the Indoor Wireless Communication (IWC) Nakagami fading channel. The up- and downlink of the system implement Differential Phase Shift Keying (DPSK) and Coherent Phase Shift Keying (CPSK) as modulation schemes respectively, and are analyzed using Maximum Entropy (MaxEnt) principles due to its reliability and accuracy. As a means to enhance system capacity and performance, different forms of diversity are investigated; for the up- and downlink, respectively, RAKE reception and Maximum Ratio Combining (MRC) diversity together with Forward Error Control (FEC) coding are assumed. Further, the validity of the Gaussian Assumption (GA) is quantified and investigated under fading and non-fading conditions by calculating the missing information, using Minimum Relative Entropy (MRE) principles between the Inter- User Interference (IUI) distribution and a Gaussian distribution of equal variance.

M.Ing. / In this thesis we investigated the characterization of wireless communication channels. The characterization of the channel itself focused on the analog as well as the discrete domains. This was done to provide a more complete treatment than that which is usually found in the literature (which normally focuses" on either one of the domains). We also provide a discussion on the basic issues that arise when trying to characterize a wireless communication channel. The analog characterization proceeded by means of the impulse response of the chan­ nel. We calculated the scattering function of the channel as well as the Delay and Doppler Power Spectra. From the wideband data we also derived the equivalent Narrowband en­ velope and phase and investigated the probability density functions of these. We also implemented the GMSK modulation scheme, used in the GSM system, in software. This included a simulation of the additive white Gaussian noise channel as well as the Rayleigh fading channel, the two most common assumed models. A complete discussion is provided on this implementation. The implementation above was used to obtain discrete channel models of GMSK on a Rayleigh fading channel. This also provided us with the opportunity to compare the performance of three of the most common discrete channel models : the Gilbert model, the Gilbert-Elliott model and the Fritchman model. Lastly we obtained equations that describe the behaviour of the Gilbert model's parameters as a function of average signal­ to-noise ratio.

No description available.

From wirelessly connected robots to car-to-car communications, and to smart cities, almost every aspect of our lives will benefit from future wireless communications. While promise an exciting future world, next-generation wireless communications impose requirements on the data rate, spectral efficiency, and latency (among others) that are higher than those for today's systems by several orders of magnitude. Full-duplex wireless, an emergent wireless communications paradigm, breaks the long-held assumption that it is impossible for a wireless device to transmit and receive simultaneously at the same frequency, and has the potential to immediately double network capacity at the physical (PHY) layer and offers many other benefits (such as reduced latency) at the higher layers. Recently, discrete-component-based demonstrations have established the feasibility of full-duplex wireless. However, the realization of integrated full duplex radios, compact radios that can fit into smartphones, is fraught with fundamental challenges. In addition, to unleash the full potential of full-duplex communication, a careful redesign of the PHY layer and the medium access control (MAC) layer using a cross-layer approach is required. The biggest challenge associated with full duplex wireless is the tremendous amount of transmitter self-interference right on top of the desired signal. In this dissertation, new self-interference-cancellation approaches at both system and circuit levels are presented, contributing towards the realization of full-duplex radios using integrated circuit technology. Specifically, these new approaches involve elimination of the noise and distortion of the cancellation circuitry, enhancing the integrated cancellation bandwidth, and performing joint radio frequency, analog, and digital cancellation to achieve cancellation with nearly one part-per-billion accuracy. In collaboration with researchers at higher layers of the stack, a cross-layer approach has been used in our full-duplex research and has allowed us to derive power allocation algorithms and to characterize rate-gain improvements for full-duplex wireless networks. To enable experimental characterization of full-duplex MAC layer algorithms, a cross-layered software-defined full-duplex radio testbed has been developed. In collaboration with researchers from the field of micro-electro-mechanical systems, we demonstrate a multi-band frequency-division duplexing system using a cavity-filter-based tunable duplexer and our integrated widely-tunable self-interference-cancelling receiver.