The growing demand for amorphous oxide semiconductor thin film transistors (TFT) necessitates the development of a physics-based field-effect mobility model that links the terminal characteristics of TFTs to their material properties. This need is particularly acute since existing approaches fail to explicitly account for the unique carrier transport properties of oxide semiconductors. The first part of this thesis specifically addresses this challenge. Here, it is shown that the electron conduction mechanism in the above-threshold regime of amorphous oxide semiconductor TFTs is controlled by both percolation and trap-limited conduction. In the limit where trap-limited conduction prevails, the characteristic temperature of tail states controls the field-effect mobility; whereas in the limit where percolation prevails, the properties of conduction’s band potential fluctuation govern it. Irrespective of the operation regime, the fieldeffect mobility is found to follow a power law form. The value of the trap density plays a critical role in determining the transition voltage. This is because a high value leads to Fermi level pinning, and thus trap-limited conduction dominates; whereas a low value results in percolation since the Fermi level becomes able to cross the conduction band edge for trap density of <1020 cm-3eV-1. The threshold voltage (VT) stability of TFTs is a critical figure-of-merit that largely determines the lifetime of active matrix backplanes. This, coupled with the growing demand on the TFTs to perform analogue functions, necessitates the study of the instability mechanisms in oxide semiconductor TFTs. The second part of this thesis specifically addresses this challenge. Here, electrical and/or optical stress/recovery measurements reveal the presence of charge trapping and persistent photoconductivity (PPC) in passivated HfInZnO/SiOx and GaInZnO/InZnO/SiOx TFT systems. In the absence of light, charge trapping is found dominant; whereas in the presence of light, PPC is. The PPC keeps the active channel in a state of high conductivity long after the removal of the light source, causing the TFT to act as though n-doped. This results in a negative VT shift, irrespective of bias magnitude and polarity. The origin of the PPC is attributed to the ionisation of oxygen vacancy sites, mainly based on the observations of its temperature- and wavelength-dependence. Interestingly, the gate voltage is found to control the decay of the PPC, giving rise to a memory action. This is explained in terms of the dependence of the rate of the PPC recovery reaction, Vo +++2e /rightarrow Vo, on carrier concentration.

Avalanches pose a significant threat to human life and settlements, hence their study is key in formulating settlement risk zones. Validating models of avalanches, developed to predict their behaviour, are limited by the quality of current field data. University College London (UCL) has developed a sophisticated FMCW phased array raday with sub-metre range resolution for imaging avalanches. The radar is installed in a bunker in an avalanche test facility in Valleé de La Sionne (VDLS) and has been gathering data since winter 2009. Due to space constraint and destructive issues with the bunker, a design for compact, low-cost and directive antenna is crucial to be used with the avalanche radar. This thesis documents the theory, analysis, design and measurement of the compact, low-cost and directive antennas. First, the parameter and location of VDLS is scrutinized to determine the characteristics that needs to be fulfilled when designing the antennas. Some of the parameters that were looked into are the operating frequency, the beamwidth of the antenna and the effects of placing the antenna in an array. Then, the study of the first antenna, a multilayer microstrip patch antenna with novel feeding method is presented. The directive concept of the antenna is obtained using an array of 4 x 2 patches, with increased antenna height and a novel feeding technique. The antenna is then fabricated using Rogers RT Duroid 5880 and assembled to be measured and compared with the simulated version. The second antenna proposed in this thesis is an array of microstrip Yagi antenna. The directive concept of this antenna is achieved through the design of the microstrip Yagi antenna itself and applying the antenna in an array of linear arrangement. The antenna is then fabricated using Rogers RT Duroid 5880 and compared with the simulated version. Finally, the first antenna characteristics is being exploited in series of Matlab program to estimate the antenna performance in a sparsely-populated phased array system. It has been demonstrated experimentally and through measurement that the antenna system design is suitable for the application of FMCW avalanche radar.

Metallic nanostructures that contain bound geometries will support localised surface plasmon (LSP) resonances if they are illuminated with light of appropriate frequency. These LSP resonances result in a concentration of the electric field of the incident light into a volume which is smaller than the photon wavelength. Certain geometries that support LSP resonances are sensitive to the polarisation of incident light, and the enhanced electromagnetic field can therefore be tuned in situ by adjusting this polarisation. We have investigated polarisation tunable LSP field enhancement by observing, in the linear regime, the interaction of an asymmetric cruciform aperture structure with a chemical bond and, in the non-linear regime, the second harmonic generation (SHG) produced by three metallic nanostructures. Numerical simulations implementing rigorous coupled-wave analysis (RCWA) were used to find asymmetric cruciform aperture dimensions that produced LSP resonances when illuminated with light of a wavelength between 2 μm and 8 μm. Arrays of these apertures were fabricated in a 35 nm thick gold film on a transparent calcium fluoride (CaF_{2}) substrate. The fabrication methods used to create the apertures were either focused ion beam (FIB) milling, or electron beam lithography (EBL) with argon ion milling, of the gold film. Fourier transform infrared spectroscopy (FTIR) was used to measure the transmission and reflection spectra of these plasmonic nanostructures. The apertures were coated with poly(methyl methacrylate) (PMMA), which has a local absorption maximum at 5.784 μm created by the stretching of its carbonyl bonds. The transmission and reflection spectra of the PMMA-coated apertures were measured using FTIR. The interaction of the LSP and molecular resonances was shown to form an asymmetric Fano resonance at the carbonyl bond wavelength. We found that this Fano resonance can be tuned in situ by rotating the polarisation of incident light. A classical mechanical oscillator model was developed to interpret the reflection and transmission spectrum in terms of the interference of the LSP and molecular resonances. A quantum mechanical model was also developed and used to predict the absorption spectrum of the system. This quantum mechanical model provides information on the physical interactions within the system, and predicts a near-field mediated interaction between the plasmon and molecular resonances. Nonlinear optical measurements were made using an SHG microscope, which allowed the location of near-field SHG hotspots to be determined. Three geometries were measured using this technique using fundamental wavelengths of 800 nm or 1 μm. The first geometry, a chiral star structure, was found to display dichroic SHG that was dependent on the handedness of the incident circularly-polarised fundamental light. The second, a `windmill' structure, was used to investigate the dependence of near-field SHG on the linear polarisation of fundamental light; the ablation of these metallic windmill structures by the fundamental demonstrates that laser ablation of patterned surfaces is dependent on the LSP resonance of the constituent structures. Finally, the spatial dependence of SHG produced by a cruciform aperture structure in a gold film illuminated by linearly polarised light was observed. SHG intensity was found to be greatest along the axis of the cruciform which was perpendicular to the incident E field polarisation.

Classification of targets is a key problem of modern radar and sonar systems. This is an activity carried out with great success by echolocating mammals, such as bats, that have evolved echolocation as a means of detecting, selecting and attacking prey over a period of more than 50 million years. Because they have developed a highly sophisticated capability on which they depend for their survival, it is likely that there is potentially a great deal that can be learnt from understanding how they use this capability and how this might be valuably applied to radar and sonar systems. Bat-pollinated plants and their flowers represent a very interesting class of organisms for the study of target classification as it is thought that co-evolution has shaped bat-pollinated flowers in order to ease classification by bats. In this thesis, the strategy that underpins classification of flowers by bats is investigated. An acoustic radar has been developed to collect data to perform a floral echoes analysis. Results show that there is a relative relevance of specific parts of the flower in displaying information to bats and show that there are different characteristics in the flowers' echo fingerprints, depending on age and stage of maturity, that bats might use to choose the most suitable flowers for pollination. We show that, as suggested by the oral echoes analysis, a more intelligent way to perform target classification can result in improved classification performance and, investigate biologically inspired methods and ideas that might become important tools for the study and the development of radar and sonar target classification.

The work described in the thesis investigates the features of control plane functionality for routing wavelength paths to serve a set of sub-wavelength demands. The work takes account of routing problems only found in physical network layers, notably analogue transmission impairments. Much work exists on routing connections for dynamic Wavelength-Routed Optical Networks (WRON) and to demonstrate their advantages over static photonic networks. However, the question of how agile the WRON should be has not been addressed quantitatively. A categorization of switching speeds is extended, and compared with the reasons for requiring network agility. The increase of effective network capacity achieved with increased agility is quantified through new simulations. It is demonstrated that this benefit only occurs within a certain window of network fill; achievement of significant gain from a more-agile network may be prevented by the operator’s chosen tolerable blocking probability. The Wavelength Path Sharing (WPS) scheme uses semi-static wavelengths to form unidirectional photonic shared buses, reducing the need for photonic agility. Making WPS more practical, novel improved routing algorithms are proposed and evaluated for both execution time and performance, offering significant benefit in speed at modest cost in efficiency. Photonic viability is the question of whether a path that the control plane can configure will work with an acceptable bit error rate (BER) despite the physical transmission impairments encountered. It is shown that, although there is no single approach that is simple, quick to execute and generally applicable at this time, under stated conditions approximations may be made to achieve a general solution that will be fast enough to enable some applications of agility. The presented algorithms, analysis of optimal network agility and viability assessment approaches can be applied in the analysis and design of future photonic control planes and network architectures.

The expansion of wireless sensor networks to advanced areas, including structure health monitoring, multimedia surveillance, and health care monitoring applications, has resulted in new and complex problems. Traditional sensor systems are designed and optimised for extremely low traffic loads. However, it has been witnessed that network performance drops rapidly with the higher traffic loads common in advanced applications. In this thesis, we examine the system characteristics and new system requirements of these advanced sensor network applications. Based on this analysis, we propose an improved architecture for wireless sensor systems to increase the network performance while maintaining compatibility with the essential WSN requirements: low power, low cost, and distributed scalability. We propose a modified architecture deriving from the IEEE 802.15.4 standard, which is shown to significantly increase the network performance in applications generating increased data loads. This is achieved by introducing the possibility of independently allocating the sub-carriers in a distributed manner. As a result, the overall efficiency of the channel contention mechanism will be increased to deliver higher throughput with lower energy consumption. Additionally, we develop the concept of increasing the data transmission efficiency by adapting the spreading code length to the wireless environment. Such a modification will not only be able to deliver higher throughput but also maintain a reliable wireless link in the harsh RF environment. Finally, we propose the use of the battery recovery effect to increase the power efficiency of the system under heavy traffic load conditions. These three innovations minimise the contention window period while maximising the capacity of the available channel, which is shown to increase network performance in terms of energy efficiency, throughput and latency. The proposed system is shown to be backwards compatible and able to satisfy both traditional and advanced applications and is particularly suitable for deployment in harsh RF environments. Experiments and analytic techniques have been described and developed to produce performance metrics for all the proposed techniques.

Field Programmable Gate Array (FPGA) devices and high-level hardware development languages represent a new and exciting addition to traditional research tools, where simulation models can be evaluated by the direct implementation of complex algorithms and processes. Signal processing functions that are based on well known and standardised mathematical operations, such as Fast Fourier Transforms (FFTs), are well suited for FPGA implementation. At UCL, research is on-going on the design, modelling and simulation of Frequency Division Multiplexing (FDM) techniques such as Spectrally E - cient Frequency Division Multiplexing (SEFDM) which, for a given data rate, require less bandwidth relative to equivalent Orthogonal Frequency Division Multiplexing (OFDM). SEFDM is based around standard mathematical functions and is an ideal candidate for FPGA implementation. The aim of the research and engineering work reported in this thesis is to design and implement a system that generates SEFDM signals for the purposes of testing and veri cation, in real communication environments. The aim is to use FPGA hardware and Digital to Analogue Converters (DACs) to generate such signals and allow recon gurability using standard interfaces and user friendly software. The thesis details the conceptualisation, design and build of an FPGA-based wireless signal generation platform. The characterisation applied to the system, using the FPGA to drive stimulus signals is reported and the thesis will include details of the FPGA encapsulation of the minimum protocol elements required for communication (of control signals) over Ethernet. Detailed testing of the hardware is reported, together with a newly designed in the loop testing methodology. Veri ed test results are also reported with full details of time and frequency results as well as full FPGA design assessment. Altogether, the thesis describes the engineering design, construction and testing of a new FPGA hardware and software system for use in communication test scenarios, controlled over Ethernet.

This project has been constituted to exceed the sampling frequencies attained in the previous Astrium satellite communication signal processors (i.e. beamforming, null steering). Thus far these systems have been synchronous as the ADCs, DACs used, had comparatively low sample rate (i.e. less than or equal to 120 MHz) and were designed to have reset capability to resynchronise the system after power up, SEUs (single event upsets), radiation and other interference effects. Ever increasing demand for higher sampling rates requires a complex clocking scheme within the ADCs, DACs and DSPs, incorporating several frequency divider circuits, bringing with them the issues of uncertainties in the clock synchronisation. The clock ambiguities can be categorised into two types: the first type is SEU or power up related bit flip (i.e. change in flip-flop or register output polarity) and the second type is related to the thermal drift and the component aging. Obviously a bit flip in a clock divider circuit will cause a clock/data cycle discrepancy on the other hand the aging and thermal drift will cause a small amount of phase delay. In order to detect the discrepancies, two different methodologies are used namely: coarse calibration for the bit flip rectification and fine calibration for aging and thermal drift errors. Once a bit flip type error is detected it is handled by a phase commutator and a plesiochronous interface combination in the DPM (Digital Processor Module) while the thermal/aging type errors are very small in magnitude (i.e. <500ps) and are corrected by directly modifying the weighting co-efficient of the phase array antenna system. The project has been divided into 4 logical sections: the section 1 is dedicated to the firmware for calibration algorithms (Chapter 3 to Chapter 5) which detects and measures the electrical length anomalies due to the SEU and power up related uncertainties as well as the thermal and aging related phase errors. The section 2 consists of hardware design (Chapters 6 to Chapter 10 – DSP Breadboard) which is intended to be used for physical testing of the calibration algorithms developed in section 1. The operational firmware design (section 3- Chapter 11), deals with the telemetry read and program parameters upload, required on board the hardware developed in section 2 (DSP Breadboard). The section 4 is the DPM (Digital Processor Module) emulator (Chapter 12 and Chapter 13) which is developed to test all the calibration functions on the Alphasat Flight Program which has inherited all the firmware and hardware developed in sections 1 to 3. Finally the overall review of the project, including the associated hardware and firmware is presented in Chapter 14.

Nanoparticle probes can unlock the potential for multimodal biomedical imaging (in vivo and in vitro) with enhanced spatial resolution and penetration depth and targeted visualization of complex organisms. This thesis demonstrates synthesis and characterization of magnetic upconversion Gd2O3 nanoparticles that can serve as bimodal probes for optical imaging in near infrared (NIR) biomedical window, where minimal tissue auto uorescence is expected, as well as magnetic resonance imaging. More speci cally, Gd2O3:Yb3+,Tm3+ and Gd2O3:Yb3+,Er3+ nanoparticles are synthesized using urea-based homogeneous precipitation method (UHP) as well as Y2O3:Yb3+/Er3+ micro and nanoparticles using conventional combustion and thermal synthesis methods. The morphological and compositional properties of nanoparticles as well as their photonic and magnetic responses are systematically analyzed to understand the role of synthesis methods and initial synthesis materials including the concentrations of Tm and Er dopants and urea solution on the properties of the synthesized particles. The upconversion nanoparticles synthesized using UHP method are spherical and monodisperse with a size distribution in the range of 60 to 150 nm and controllable dopant concentration through manipulation of initial synthesis chemistry. When excited with 975 nm NIR radiation, Gd2O3:Yb3+,Tm3+ nanoparticles show a pure near infrared emission centered at around 810 nm (i.e., NIR-to-NIR upconversion) in NIR biological window with potential for high depth optical imaging while Er3+ doped particles emit light mainly in visible red centred at around 661 nm. The photoluminescence and transient optical decay measurements demonstrate distinctly di erent energy transfer mechanisms for Er and Tm doped samples. While these measurements signify a dominant role for Yb3+ dopants in strong upconversion emission of Tm3+ samples with a double exponential decay behaviour, they show less important role of Yb3+ in Er3+ samples with a single exponential decay. Systematic magnetic characterization demonstrate strong paramagnetic behaviour for the optically active upconversion nanoparticles, demonstrating their potential for bimodal optical and magnetic resonant imaging.

A well known problem in peer-to-peer overlays is that no single entity has control over the software, hardware and configuration of peers. Thus, each peer can selfishly adapt its behaviour to maximise its benefit from the overlay. This thesis is concerned with the modelling and design of incentive mechanisms for QoS-overlays: resource allocation protocols that provide strategic peers with participation incentives, while at the same time optimising the performance of the peer-to-peer distribution overlay. The contributions of this thesis are as follows. First, we present PledgeRoute, a novel contribution accounting system that can be used, along with a set of reciprocity policies, as an incentive mechanism to encourage peers to contribute resources even when users are not actively consuming overlay services. This mechanism uses a decentralised credit network, is resilient to sybil attacks, and allows peers to achieve time and space deferred contribution reciprocity. Then, we present a novel, QoS-aware resource allocation model based on Vickrey auctions that uses PledgeRoute as a substrate. It acts as an incentive mechanism by providing efficient overlay construction, while at the same time allocating increasing service quality to those peers that contribute more to the network. The model is then applied to lagsensitive chunk swarming, and some of its properties are explored for different peer delay distributions. When considering QoS overlays deployed over the best-effort Internet, the quality received by a client cannot be adjudicated completely to either its serving peer or the intervening network between them. By drawing parallels between this situation and well-known hidden action situations in microeconomics, we propose a novel scheme to ensure adherence to advertised QoS levels. We then apply it to delay-sensitive chunk distribution overlays and present the optimal contract payments required, along with a method for QoS contract enforcement through reciprocative strategies. We also present a probabilistic model for application-layer delay as a function of the prevailing network conditions. Finally, we address the incentives of managed overlays, and the prediction of their behaviour. We propose two novel models of multihoming managed overlay incentives in which overlays can freely allocate their traffic flows between different ISPs. One is obtained by optimising an overlay utility function with desired properties, while the other is designed for data-driven least-squares fitting of the cross elasticity of demand. This last model is then used to solve for ISP profit maximisation.