In this work the analysis of hot carrier generation and injection in insulated-gate field effect transistors (IGFETs) is carried out with the aid of offset gate transistors. Several sets of transistors with systematically incremented gate offsets were produced for this investigation. The inclusion of photolithographically and angled-implant defined LDD (lightly doped drain) regions was also attempted, but no consistent calibration technique was found for the former and fabrication problems made the latter impractical. The charge pumping technique and the gated diode current method were employed in the evaluation of the generation of interface states and fixed charges. An adapted double-pulse charge pumping method was used to assess the energy distribution of interface states across the forbidden zone. The use of offset gate transistors made it possible to identify hot electron and hot hole degradation separately under different biasing conditions. Additionally, it suggested a clearer interpretation of the possible mechanisms involved in the degradation process.

The main purpose of this research was to obtain numerical and experimental information that may support Radar survey design. The reference material is usually concrete, but the results can be extended to other materials. Other techniques currently being employed either on concrete or masonry are also critically described. Justification for this research into Radar is clearly identified. A numerical study was carried out giving emphasis to Radar resolution (horizontal and vertical) and clutter (signal scattered from material heterogeneity). Assumptions are made for the centre frequency of transmitted pulses, and the results are compared to experiments. The experimental work focuses on signal distortion due to antenna coupling and wave attenuation during propagation in lossy media. The centre frequency of the recorded signals is used as a reference for comparisons. Digital signal processing techniques are used to analyse the experimental data, in both time domain and frequency domain. Other important aspects that could influence the accuracy of travel time measurements are also discussed, such as the antenna size for short ranges and the change in pulse shape. Radar confirms its potential as a non-destructive technique, however users have to be aware of its limitations. Experiments suggest that manufactures of Radar need to improve their systems, particularly for their portability and antenna efficiency.

This thesis analyses in detail the technology required for wireless power transfer via radio frequency (RF) ambient energy harvesting and an inductive power transfer system (IPT). Radio frequency harvesting circuits have been demonstrated for more than fifty years, but only a few have been able to harvest energy from freely available ambient (i.e. non-dedicated) RF sources. To explore the potential for ambient RF energy harvesting, a city-wide RF spectral survey was undertaken in London. Using the results from this survey, various harvesters were designed to cover four frequency bands from the largest RF contributors within the ultra-high frequency (0.3 to 3 GHz) part of the frequency spectrum. Prototypes were designed, fabricated and tested for each band and proved that approximately half of the London Underground stations were found to be suitable locations for harvesting ambient RF energy using the prototypes. Inductive Power Transfer systems for transmitting tens to hundreds of watts have been reported for almost a decade. Most of the work has concentrated on the optimization of the link efficiency and have not taken into account the efficiency of the driver and rectifier. Class-E amplifiers and rectifiers have been identified as ideal drivers for IPT applications, but their power handling capability at tens of MHz has been a crucial limiting factor, since the load and inductor characteristics are set by the requirements of the resonant inductive system. The frequency limitation of the driver restricts the unloaded Q-factor of the coils and thus the link efficiency. The system presented in this work alleviates the use of heavy and expensive field-shaping techniques by presenting an efficient IPT system capable of transmitting energy with high dc-to-load efficiencies at 6 MHz across a distance of 30 cm.

In many distant speech acquisition scenarios, such as hands-free telephony or teleconferencing, the desired speech signal is corrupted by noise and reverberation. This degrades both the speech quality and intelligibility, making communication difficult or even impossible. Speech enhancement techniques seek to mitigate these effects and extract the desired speech signal. This objective is commonly achieved through the use of microphone arrays, which take advantage of the spatial properties of the sound field in order to reduce noise and reverberation. Spherical microphone arrays, where the microphones are arranged in a spherical configuration, usually mounted on a rigid baffle, are able to analyze the sound field in three dimensions; the captured sound field can then be efficiently described in the spherical harmonic domain (SHD). In this thesis, a number of novel spherical array processing algorithms are proposed, based in the SHD. In order to comprehensively evaluate these algorithms under a variety of conditions, a method is developed for simulating the acoustic impulse responses between a sound source and microphones positioned on a rigid spherical array placed in a reverberant environment. The performance of speech enhancement algorithms can often be improved by taking advantage of additional a priori information, obtained by estimating various acoustic parameters. Methods for estimating two such parameters, the direction of arrival (DOA) of a source (static or moving) and the signal-to-diffuse energy ratio, are introduced. Finally, the signals received by a microphone array can be filtered and summed by a beamformer. A tradeoff beamformer is proposed, which achieves a balance between speech distortion and noise reduction. The beamformer weights depend on the noise statistics, which cannot be directly observed and must be estimated. An estimation algorithm is developed for this purpose, exploiting the DOA estimates previously obtained to differentiate between desired and interfering coherent sources.

Random Forest (RF) is one of the most widely used supervised learning methods available. An RF is ensemble of decision tree classifiers with injection of several sources of randomness. It demonstrates a set of improvement over single decision and regression trees and is comparable or superior to major classification tools such as support vector machine (SVM) and adaptive boosting (Adaboost) with respect to accuracy, interpretability, robustness and processing speed. RF can be generally divided into training process and predicting process. Recently with emergence of large-scale data mining applications, the RF training process implemented in software on a single computer can no longer induce a complex RF model within reasonable amount of time. Alternative solutions involving computer clusters and GPUs usually come with disadvantages with respect to Performance/Power ratio and are not feasible for portable/embedded applications. In this work a set of FPGA-based implementations of the RF training process are proposed. FPGA devices allow construction of efficient custom hardware architectures and feature lower power consumption than typical GPPs or GPUs therefore are suitable for portable/embedded applications. The proposed hardware training architectures take advantage of different types of inherent parallelism in the RF training algorithm and distribute the workload to a set of parallel workers. Combining the parallel processing techniques with custom hardware designs featuring low latency, the architectures are able to accelerate the training process without loss in accuracy.

In this study, we mainly characterize the wearable antenna system for off-body communications with respect to two conventional wireless communication systems as, cellular mobile systems (pes, GSM and UMTS) and wireless local area networks (WLAN). Unlike antennas embedded in portable devices, the complicated bodycentric environment has emerged with special requirements for wearable antennas design, like compact dimensions, light weight and flexible structure, hidden or water proofmg, and most importantly, capable of providing certain radiation shielding into the human body. This thesis aims to fmd an optimum solution to meet the particular requirements ofwearable antenna design. The first part is primarily concerned with characterising the electromagnetic properties of some textile and leather materials. Both insulating and conducting materials are investigated for using as substrate and radiating elements of high perfonnance textile antennas. Then a few of new antenna designs are proposed in the second part. These antennas are made out of textile and leather materials. They are low profIle, planar in geometry, and most importantly, they are capable to provide multi-operations with considerable wide bandwidth. An electromagnetic band gap structure is studied as it can provide a high impedance ground plane for low profIle antennas. The EBG plane can provide a perfect shielding layer for the body, and reduce the radiation toward the body significantly. Furthermore, the EBG plane is able to reduce the detuning of the antenna when placing near the body without serious bandwidth reduction, increase antenna gain and reduce mutual coupling from other devices. The final part carries out a series ofexperiments which can represent the body-centric environment These include bending, washing, wearing, coupling and SAR investigations. Antenna and EBG performances under body worn environment are tested and discussed in this thesis.

The main topics of this research are modelling, fault monitoring, and cross-directional control of a plastic film manufacturing process operated by DuPont Teijin Films Ltd. The developed model is of high dimension and built using the first-principles of chemical and mechanical engineering, such as equations for mass transfer, heat transfer, and the flow of viscous fluids in addition to empirical knowledge related to the behaviour of polymer. The model in turn provides a safe off-line platform for developing new cross-directional control and fault monitoring systems. As with other sheet-forming processes, such as papermaking and steel rolling, the plastic film manufacturing process employs large arrays of actuators spread across a continuously moving sheet to control the cross-directional profiles of key product properties. In plastic manufacturing, the main control property is finished product thickness profile as measured by a scanning gauge downstream from the actuators. The role of the cross-directional control system is to maintain the measured cross-directional profiles of plastic properties on target. The second part of this research develops a novel cross-directional controller, which is in turn demonstrated by application to the first-principles model. Fault monitoring systems can be broadly classified into 3 categories: model-based, data-driven, and knowledge-based. The third part of this research introduces a novel model-based fault monitoring system. The system is demonstrated by application to both the first-principles model and industrial data extracted from the real-life plant.

Wind power has become established as an alternative power source that forms a significant proportion of national energy generation. An increasing proportion of turbines is being constructed offshore to exploit higher average wind speeds and to avoid development issues associated with onshore wind farms. Isolated locations and unpredictable weather conditions lead to increased access costs for operators when conducting scheduled and unscheduled maintenance and repairs. This has increased interest in condition monitoring systems which can track the current state of components within a wind turbine and provide operators with predicted future trends. Asset management can be improved through condition based maintenance regimes and preventative repairs. Development of novel condition monitoring systems that can accurately predict incipient damage can optimise operational performance and reduce the overall level of wind turbine generation costs. The work described in this thesis presents the development of novel sensors that may be applied to monitor wind turbine gearboxes, a component that experiences relatively high failure rates and causes considerable turbine downtime. Current systems and technology that may be adapted for use in wind turbine condition monitoring are evaluated. Lubrication related monitoring systems have been identified as an area that could be improved and are divided into those that track liberated wear material suspended in the lubricant and those that assess the state of the lubricant itself. This study presents two novel lubrication based gearbox monitoring sensors that potentially offer a low cost solution for continuous data capture. The first demonstrates the potential for active pixel sensors such as those found in digital cameras to capture images of wear particles within gearbox lubricants. Particle morphology was tracked in this system, allowing the type of particles to be correlated with the type of wear that is generated and a potential source. The second sensor uses a targeted form of infra-red absorption spectroscopy to track changes in the lubricant chemistry due to the increase in acidity. Ensuring the lubricant is functioning correctly decreases component stress and fatigue, reducing maintenance requirements.

This work considers current sourced powerelectronic converters. The thesis classifies and presents several new single-phase and three-phase differential-mode current source inverters that evolve from the basic dc-dc converter topologies. The switched, large-signal, and small-signal models of these converters are presented, and used to develop control strategies for the proposed differential-mode inverters, considering their inversion and rectification modes. The viability of each differential mode inverter/rectifier is validated using simulations and experimentation. The performances of different proposed buck, boost, and buck-boost current source inverters are discussed and compared in terms of efficiency, total harmonic distortion, input current ripple, capacitor stresses, and control complexity. Some of the proposed current source inverters offer buck-boost capability (can operate with output voltage less or greater than the input dc voltage), which is suited for grid-connected operation of single-stage three-phase buck-boost inverters. Phase variables and synchronous frame controllers are used to provide satisfactory inverter operation in inversion and rectification modes. The inherent low-order harmonic currents in the input and output of the proposed converters (predominantly, negative sequence 2nd order harmonic) are supressed using PI and PR controllers. Also, interleaved carriers are used to reduce the input current ripple of the three-phase inverters. The proposed converters can operate over a full control range from 0 to unity power factor, with power flow in both directions (unlike the conventional six-pulse current inverter). Additionally, a nonlinear control strategy, sliding mode control, is implemented with to achieve faster dynamic inverter response during faults as well as elimination of dccurrent injection into ac grid. This is necessary during unbalanced operation. Operation of single-phase differential-mode buck-boost inverters is presented, including suppression of the 2nd order harmonic in the input dc current by two methods. In the first, active suppression of the 2nd harmonic uses a power electronic circuit. The second method manipulates the modulating signal in combination with a relatively large capacitor to trap the oscillating power that causes the 2nd order harmonic to appear input dc link current. The two single-phase harmonic suppression approaches are compared.

The Discrete Cosine Transform(DCT) is the basis for current video standards like H.261, JPEG and MPEG. Since the DCT involves matrix multiplication, it is a very computationally intensive operation. Matrix multiplication entails repetitive sum of products which are carried out numerous times during the DCT computation. Therefore, as a result of the multiplications, a significant amount of switching activity takes place during the DCT process. This thesis proposes a number of new implementation schemes that reduce the switching capacitance within a DCT processor for either JPEG or MPEG environment. A number of generic schemes for low power VLSI implementation of the DCT are presented in this thesis. The schemes target reducing the effective switched capacitance within the datapath section of a DCT processor. Switched capacitance is reduced through manipulation and exploitation of correlation in pixel and cosine coefficients during the computation of the DCT coefficients. The first scheme concurrently processes blocks of cosine coefficient and pixel values during the multiplication procedure, with the aim of reducing the total switched capacitance within the multiplier circuit. The coefficients are presented to the multiplier inputs as a sequence, ordered according to bit correlation between successive cosine coefficients. The ordering of the cosine coefficients is applied to the columns. Hence the scheme is referred to as <i>column-based</i> processing. Column-Based processing exhibits power reductions of up to 50% within the multiplier unit. Another scheme, termed <i>order-based</i>, is based on the ordering of the cosine coefficients based on row segments. The scheme also utilises bit correlation between successive cosine coefficients. The effectiveness of this scheme is reflected in a power savings of up to 245. The final scheme is based on manipulating data representation of the cosine coefficients, through cosine word coding, in order to facilitate for a shift-only computational process. This eliminates the need for the multiplier unit, which poses a significant overhead in terms of power consumption, in the processing element. A maximum power saving of 41% was achieved with this implementation.