Terahertz (THz) wireless communication systems over 300 GHz can offer broad bandwidth, but they have limitations, such as high precision fabrication requirements, high fabrication cost and lossy system components. This thesis introduces a THz communication system concept based on low-cost and high precision micromachining techniques, also a new design approach to achieve a low loss system for THz communication components and systems. In order to minimise the components and systems losses, this thesis proposes a design approach for filtering amplifiers which removes the lossy planar matching and interconnection circuits to complete the transistor amplifiers design. The technique also adds filters resulting in a combined filtering amplifier. To enable this, an analytical design method for filtering amplifiers based on the conventional passive coupling matrix for filters is expanded for the filtering amplifiers. Novel coupled resonator circuits, including a general transistor model for the first time is investigated and a new [\(T\)] matrix is introduced to facilitate the transistor element. Complete mathematical formulas are provided in this thesis for the coupled resonator circuits for transistor amplifiers. Two physical design examples at X band are provided in this thesis to demonstrate the usage of the novel coupling matrix with a simple mathematical synthesis. A waveguide filtering amplifier is demonstrated using waveguide resonators. This waveguide amplifier reflects the design considerations and methods for future THz amplifier developments. Good correlations between mathematical synthesis results and physical measurements are achieved for the waveguide filtering amplifier, with 3 clearly observed poles in passband, verifying the order of the filters.

The principle of the multi-scale medial axis (MMA) is important in that any object is detected at a blurring scale proportional to the size of the object. Thus it provides a sound balance between noise removal and preserving detail. The robustness of the MMA has been reflected in many existing applications in object segmentation, recognition, description and registration. This thesis aims to improve the computational aspects of the MMA. The MMA is obtained by computing ridges in a “medialness” scale-space derived from an image. In computing the medialness scale-space, we propose an edge-free medialness algorithm, the Concordance-based Medial Axis Transform (CMAT). It not only depends on the symmetry of the positions of boundaries, but also is related to the symmetry of the intensity contrasts at boundaries. Therefore it excludes spurious MMA branches arising from isolated boundaries. In addition, the localisation accuracy for the position and width of an object, as well as the robustness under noisy conditions, is preserved in the CMAT. In computing ridges in the medialness space, we propose the sliding window algorithm for extracting locally optimal scale ridges. It is simple and efficient in that it can readily separate the scale dimension from the search space but avoids the difficult task of constructing surfaces of connected maxima. It can extract a complete set of MMA for interfering objects in scale-space, e.g. embedded or adjacent objects. These algorithms are evaluated using a quantitative study of their performance for 1-D signals and qualitative testing on 2-D images.

The world-wide application of Phasor Measurement Units (PMUs) brings great benefit to power system state estimation. The synchronised measurements from PMUs can increase estimation accuracy, synchronise states among different systems, and provide greater applicability of state estimation in the transient condition. However, the integration of synchronised measurements with state estimation can introduce efficiency problems due to the substantial burden of data. The research is divided into two parts: finding a solution to cope with the computational efficiency problem and developing a transient state estimation algorithm based on synchronised measurements from PMUs. The computational efficiency problems constitute important considerations in the operation of state estimation. To improve the low computational efficiency, two distributed algorithms are proposed in Chapters 4 and 5. In these two algorithms, the modelling, structure, and solution are described, and the corresponding procedures of bad data processing are presented. Numerical results on the IEEE 30-bus, 118-bus and 300-bus systems can verify the effectiveness of the two proposed algorithms. A novel transient state estimation algorithm based on synchronised measurements is proposed in Chapter 6. Considering the scanning cycle and sampling rate of PMU measurements, the proposed algorithm can estimate transient states in a practical way. The performance of the proposed algorithm is demonstrated in a transient simulation on the IEEE 14-bus system.

When humans summarize a document they usually read the text first, understand it then attempt to write a summary. In essence, these processes require at least some basic level of background knowledge by the reader. The least of which would be the Natural Language the text is written in. In this thesis, an attempt is made to bridge the gap of machines understanding by proposing a framework backed with knowledge repositories constructed by humans and containing real human concepts. I use WordNet, a hierarchically-structured repository that was created by linguistic experts and is rich in its explicitly defined lexical relations. With WordNet, algorithms for computing the semantic similarity between terms were proposed and implemented. These algorithms were especially useful when applied to the application of Automatic Documents Summarization as shown with the obtained evaluation results. I also use Wikipedia, the largest encyclopedia to date. Because of its openness and structure, three problems had to be handled in this thesis: Extracting knowledge and features from Wikipedia, enriching the representation of text documents with the extracted features, and using them in the application of Automatic Summarization. When applying the features extractor to a summarization system, competitive evaluation results were obtained.

Ziel der Arbeit ist die Erweiterung des Software-Werkzeuges ¨SAPCDEM¨ um eine Skriptsprache. Dieses Werkzeug wurde in der Abteilung ¨SAP R/3 Software Factory¨ der SAP AG entwickelt und dient der Verwaltung und Qualitaetssicherung der SAP R/3 CD's. Im Rahmen dieser Arbeit wird das Werkzeug hinsichtlich seines Aufbaus und seiner Arbeitsweise analysiert. Es werden die genauen Anforderungen an die Skriptsprache ermittelt und Loesungsansaetze diskutiert, diese in das vorhandene System zu integrieren. Ein Interpreter wird entworfen, der die Anweisungen der Sprache in entsprechende Aktionen des Werkzeuges umsetzt.

As devices are scaled to gate lengths of sub 100 nm the effects of intrinsic parameter fluctuations will become increasingly important. This work presents a systematic simulation study of intrinsic parameter fluctuations, consisting of random dopant fluctations, line edge roughness and oxide thickness fluctuations, in a real 35 nm MOSFET developed by Toshiba. The simulations are calibrated against experimental data for the real device and it is found that discrete random dopants have the greatest impact on both the threshold voltage and leakage current fluctuations with a σVT of 33.2mV and a percentage increase in the average leakage current of 50%. Line edge roughness has the second greatest impact with a σVT of 19mV and percentage increase in the average leakage current of 45.5%. The smallest impact is caused by oxide thickness variations resulting in a σVT of 1.8mV and a 13% increase in the average leakage current. The combined effects of pairs of fluctuations is also studied, showing that these sources of intrinsic parameter fluctuations are statistically independent and a calculated σVT of 39mV is given for all of the sources combined. This value is on par with that reported in literature for the 90 nm technology node.

The work reported In this thesis is concerned with the fabrication and characterisation of microwave lumped element overlay capacitors for used at 10 GHz. The research has included the investigation and selection of suitable materials for use in the component manufacture, and,the development and optimisation of the necessary fabrication processes, which at the outset of' the project, were not available to the author. For this reason, and for the benefit of other research workers involved in the construction of microwave thin film components, the presentation of the processing data in this thesis is biased towards step-by-step accounts of' each of the major processes. The processing data is then readily available in a practical, and useful, form. Of' particular interest in the processing field, is the novel technique developed for the deposition of' the capacitor dielectric material without etching. This work was carried out in collaboration with a fellow research student, and is covered by a U.K. Patent. The microwave measurement of the capacitor properties have been performed using microstrip resonators. To enable these measurementsresonators to be accurately designed, measurements have been performed on, and results documented for, the transmission characteristics of microstrip lines on quartz substrates, and for the properties of micros trip gap discontinuities, and microstrip gap-and-step discontinuities, at X band. The technique for the measurement of the overlay capacitor properties is novel in that the capacitor is measured "in situ", and bond wires, which are a characteristic of' other measurement techniques, are not required. This permits a higher measurement accuracy, due to the reduction in connector parasitics. In addition, the equivalent circuit approach to the calculation of the capacitor properties is used, and not the perturbation approach, which permits large changes in frequency and Q factor during the course of the measurements, and this allows very measurements to be obtained. Results are presented for measurements performed on overlay capacitors, using silicon dioxide and alumina dielectric materials, and these indicate that capacitors with useful values of capacitance and Q factor can be produced at X band. Theories are presented for the capacitor electrode inductance and resistance, and these enable the dielectric constant and Q :factor of the capacitor dielectric materials, to be evaluated from the measured ccapacitor properties. Values of' the dielectric properties thus obtained are shown to be in good agreement with the characteristics of the dielectric materials in isolation, measured by a :fellow research student, using cavity techniques.

The thesis details the development of a short range, multi-hop underwater optical wireless sensor network. Multi-hop underwater optical wireless communication using a line of sight (LOS) link can provide a greater range compared to a single hop network, and provide physically secure connections for underwater sensor networks. This kind of system can be very power efficient, and supported data rate can be from tens of kbps up to a few hundred kbps. The aims were to build a cheap communication prototype using “off the shelf” components, such as a microcontroller, optoelectronics etc. for demonstration purpose. To support the built prototype, a directional MAC protocol has been developed which considers the directionality of light propagation. The multi-hop approach has not been considered for underwater optical wireless communication before, while most of the research focus is to develop long range and high powered communication links. In this thesis, a custom built transceiver using blue and green LEDs has been developed, which supports a data rate up to 140kbps, when the NRZ-OOK modulation technique is used. For the transmitter part, a digital LED driver has been used, while on the receiver side, a transimpedance amplifier using a single transistor has been developed. This configuration for optical wireless receiver system design has not been usual, but it works very well for the proposed prototype. A second stage voltage amplifier was also designed to boost the signal up to 5V for the microcontroller, which was also based on transistors. To demonstrate the principle of multi-hop communication, a line-type network prototype using two sensor nodes and a gateway node has been designed, built and tested in the lab environment. Each node was equipped with two transceivers controlled by a microcontroller to make a full-duplex communication system. To minimize the cost, all components of a node were built on a single PCB board. To upload data from the sensor node to the gateway node, a green LED has been used, and to transmit the control signal from gateway node to sensor nodes, a blue LED was used. For the demonstration purpose the communication range was considered up to 1m, which can be increased significantly by using high powered LED, and external optics such as lenses, concentrators, etc. A directional MAC protocol has been designed, considering the directionality of the network. The designed protocol is based on TDMA techniques, but modified for the proposed application. The gateway node controls all other nodes in the network and acts as a master node. Because of the directional full-duplex network, there is much less chance of a collision, when using a TDMA approach. Therefore, a random access protocol was not needed for the proposed architecture. Finally, experimental results validate the fact that the multi-hop approach is a viable solution to increase the communication ranges for underwater optical wireless sensor networks. Different sets of experiments show that the proposed system can be implemented in the real environment, such as, oceans, canals and ponds.

The objective of this thesis is to develop an intelligent diagnostic technique POET (Passive Optical Emission Tomography) for the investigation of in cylinder combustion chemiluminescence. As a non-intrusive optical system, the POET system employs 40 fibre optic cables connected to 40 PMTs (Photo Multiplier Tube) to monitor the combustion process and flame front propagation in a modified commercial OHV (Over Head Valve) Pro 206 IC engine. The POET approach overcomes several limitations of present combustion research methods using a combination of fibre optic detection probes, photomultipliers and a tomographic diagnostics. The fibre optic probes are placed on a specially designed cylinder head gasket for non-invasively inserting cylinder. Each independent probe can measure the turbulent chemiluminescence of combustion flame front at up to 20 kHz. The resultant intensities can then be gathered tomographically using MART (Multiplicative Algebraic Reconstruction Technique) software to reconstruct an image of the complete flame-front. The approach is essentially a lensless imaging technique, which has the advantage of not requiring a specialized engine construction with conventional viewing ports to visualize the combustion image. The fibre optic system, through the use of 40, 2m long thermally isolated fibre optic cables can withstand combustion temperatures and is immune from electronic noise, typically generated by the spark plug. The POET system uses a MART tomographic methodology to reconstruct the turbulent combustion process. The data collected has been reconstructed to produce a temporal and spatial image of the combustion flame front. The variations of lame turbulence are monitored by sequences of reconstructed images. Therefore, the POET diagnostic technique reduces the complications of classic flame front propagation measurement systems and successfully demonstrates the in-cylinder combustion process. In this thesis, a series of calibration exercises have been performed to ensure that the photomultipliers of the POET system have sufficient temporal and spatial resolution to quantitatively map the flow velocity turbulence and chemiluminescence of the flame front. In the results, the flame has been analyzed using UV filters and blue filters to monitor the modified natural gas fuel engine. The flame front propagation speed has been evaluated and it is, on average, 12 m/s at 2280 rpm. Sequences of images have been used to illustrate the combustion explosion process at different rpm.

Liquid sensors are required for a multitude of applications in the food and beverage sectors, in the pharmaceutical industry or environmental monitoring. The focus of this work is on the development of high frequency shear horizontal surface acoustic wave (SH-SAW) sensors for liquid media identification and characterisation. Among the various types of surface acoustic wave modes propagating in solids, the SH-SAWs were found to be the most suitable for operation in liquids. Dual delay line and resonator sensor configurations were designed and fabricated on lithium tantalate (LiTa03) substrates; the design and the subsequent fabrication procedures of the sensors are described in detail. Furthermore, the electrical characterization of the sensors was carried out with a network analyser, and a comparative analysis was performed between sensors with different configurations. The interdigital transducers, used as the interface between the electrical and acoustic domains, presented good reflection coefficients and had near perfect matched impedances and return loss figures up to 45 dB. The insertion loss of the sensors varied with the surface conditions while it was improved by using total or partial metallization of the surface or employing grating structures on the propagation path. The SH-SAW devices were exposed to basic taste solutions and all the sensor configurations tested were able to discriminate them well. Measurements were done in both standard wired set-ups and a semi-wireless set-up, thus proving the sensor's capability for remote operation. Further investigations regarding the electronic tongue applicability of the SH-SAW sensors were conducted on a two port resonator device. The resonator was tested with six basic taste solutions, with taste solutions with varying concentrations, with binary mixtures of taste solutions and proved successful in identifying all test samples. A multivariate analysis was performed on the resonator data, and confirmed that the sensor's responses are influenced by the physical properties of the tested solutions. The multiple linear models derived are statistically significant and can explain high percentage of the data variability, offering a simplified alternative to the complex analytical models of the SH-SAW sensors. Also, a voltage modulated sensor system was proposed for smart assaying of biomaterials and its operation principle is described. The preliminary tests carried out showed a significant voltage effect on carbon nanoparticles. The voltage modulated system is proposed as an analytical microsystem for the screening of bacterial cells. All sensors in this project had no bio-chemical selective layer making them nonspecific, yet they create robust, durable and low-cost systems.