Multiple-input multiple output (MIMO) systems deploy multiple antennas at either end of a communication link and can provide significant benefits compared to traditional single antenna systems, such as increased data rates through spatial multiplexing gain, and/or improved link reliability through diversity techniques. Recently, the natural extension of utilising multiple antennas in relay networks, known as MIMO relaying, has attracted significant research attention due to the fact that the benefits of MIMO can be coupled with extended network coverage through the use of relaying devices. This thesis concentrates on the design and analysis of different aspects of MIMO relay systems communicating over frequency selective channels with the use of orthogonal frequency division multiplexing (OFDM). The first focus of this thesis is on the development of training based channel estimation algorithms for two-hop MIMO OFDM relaying. In the first phase of channel estimation the relay-dest ination channel is estimated using conventional point-to-point MIMO estimation techniques. In the second phase, the source sends known training symbols to the relay, which precodes the received symbols and forwards them to the destination. In order to estimate the source-relay channel at the destination, an iterative algorithm is derived, which involves sequentially solving a number of convex optimisation problems and has guaranteed convergence. Since the proposed iterative algorithm may be too computationally complex for practical systems, a simplified approach is also derived where the channel estimation processors can be calculated in closed form. Under the assumption of perfect channel state information (CSI), we then develop non-linear transceiver designs for MIMO OFDM relay systems, focusing specifically on decision feedback equalisation (DFE) and Tomlinson Harashima precoding (THP). The optimal source and relay precoding matrices are derived that minimise the arithmetic mean square error (MSE) subject to source and relay transmission power constraints, when either a zero forcing (ZF) or minimum mean square error (MMSE) equaliser is used at the destination. Simulation results demonstrate that the proposed non-linear solutions outperform linear transceivers in terms of bit error rate (BER) and MSE. For the case of imperfect CSI at all nodes, robust DFE and THP transceivers are then considered that aim to minimise the expected artithmetic MSE subject to the source and relay transmission power constraints. The channel estimation errors are modelled as being drawn from matrix variate Gaussian distributions with known mean and covariance. The source and relay precoder structures are derived for the case that the optimal MMSE equaliser is used at the destination. The derived precoder structures are shown to be optimal for the special case that the channel estimation errors are uncorrelated. Simulation results demonstrate the robustness of the proposed algorit hms to channel estimation errors.

Implementation of distributed energy resources (DER) has the potential to lower the carbon oxygen emissions, reduce the power distribution losses and improve the overall system operation. Despite the numerous advantages brought by these small-scale DERs, effective protection and control of such systems are still unsolved challenges. The distribution system is increasingly being confronted with congestion and voltage problems, which limits further penetration of DERs. Numerous studies have been conducted to analyse these challenges and provide recommendations or guidance for protection and control in the past few years. There is also a lot of effort to develop an advanced regime for integration of large amounts of DERs, such as the "microgrid". Microgrids are designed to provide control and protection of a cluster of DERs, storage units and loads in a way that can coordinate with the conventional utility grid operation with little conflict. As flexible as it is, a microgrid is connec ted to the utility grid behaving as a controllable entity in normal operation, and can be disconnected from the grid to present itself as a power island in emergency, e.g. system black-out. Since most of the DERs are interfaced with inverters, this thesis is dedicated to provide in-depth investigation of protection and control within inverter dominated microgrids. The thesis provides two main valuable contributions. Firstly, an enhanced control scheme for a microgrid consisting of multi inverter interfaced generators (IIG) is developed and compared to the conventional droop based decentralized control. The proposed control scheme is particularly designed for systems with IIGs interconnected via relatively long cable lengths (several kilo metres). It also allows switchless mode transition between islanded operation and grid-connected operation, which reduces the transient voltage and current oscillations, and enhances the transient behaviour of the IIGs. Compared to the conventional droop based decentralized control, the proposed control scheme has better operational stability and is immune to different lengths and R/X ratios of connecting cables. The proposed control also brings better voltage regulation and has a larger power output capacity. Secondly, a new travelling wave based protection scheme is developed which involves modification of an application friendly signal processing technique - Mathematical Morphology. The impact of distance to fault, fault inception angle and fault impedance is analysed and quantified. The thesis proposes a systematic protection solution which is proved to be immune to the changes of system topology, modes of operation and load conditions.

Fatigue is a common symptom of weakness either physically or mentally. These symptoms may led to a drop in motivation, weakened sensitivity, slowing of responsiveness and inability to give full attention. All of these problems can cause adverse effects, such as accidents, especially those that require full attention as drivers of vehicles, and rail operators, the pilot of an aircraft or ship operators. This research investigates systems to detect and quantify the signs of fatigue using non-invasive facial analytics. There are four main algorithms that represent the major contribution from the PhD research. These algorithms encompass facial fatigue detection and quantification system as a whole. Firstly, a new technique to detect the face is introduced. This face detection algorithm is an affiliation of colour skin segmentation technique, connected component of binary image usage, and learning machine algorithm. The introduced face detection algorithm is able to reduce the false pos itive detection rate by a very significant margin. For the facial fatigue detection and quantification, the major fatigue signs features are from the eye activity. A new algorithm called the Interdependence and Adaptive Scale Mean Shift (IASMS) is presented. The IASMS is able to quantify the state of eye as well as to track non-rigid eye movement. IASMS integrates the mean shift tracking algorithm with an adaptive scale scheme, which is used to track the iris and quantify the iris size. The IASMS is associated with face detection algorithm, image enhanced scheme, eye open detection technique and iris detection method in the initialisation process. This proposed method is able to quantify the eye activities that represent the blink rate and the duration of eye closure. The third contribution is yawning analysis algorithm. Commonly yawning is detected based on a wide mouth opening. Frequently however this approach is thwarted by the common human reaction to hand-cover the mouth during yawning. In this research, a new approach to analyse yawning which takes into account the covered mouth is introduced. This algorithm combines with a new technique of mouth opening measurements, covered mouth detection, and facial distortion (wrinkles) detection. By using this proposed method, yawning is still able to detect even though the mouth is covered. In order to have reliable results from the testing and evaluating of the developed fatigue detection algorithm, the real signs of fatigue are required. This research develops a recorded face activities database of the people that experience fatigue. This fatigue database is called as the Strathclyde Fatigue Facial (SFF). To induce the fatigue signs, ethically approved sleep deprivation experiments were carried out. In these experiments twenty participants, and four sessions were undertaken, which the participant has to deprive their sleep in 0, 3, 5, and 8 hours. The participants were subsequently requested to carry out 5 cognitiv e tasks that are related to the sleep loss. The last contribution of this research is a technique to recognise the fatigue signs. The existing fatigue detection system is based on single classification. However, this work presents a new approach for fatigue recognition which the fatigue is classified into levels. The levels of fatigue are justified based on the sleep deprivation stages where the SFF database is fully used for training, testing and evaluation of the developed fatigue recognition algorithm. This fatigue recognition algorithm is then integrated into a Fatigue Monitoring Tool (FMT) platform. This FMT has been used to test the participant that carried out the tasks as ship crew in shipping bridge simulator.

Power companies around the world are incorporating wind power into their electricity networks. Wind power is an intermittent source of energy and its technical and financial impacts on the transmission and distribution networks are not yet totally investigated. This thesis investigates the impact of wind power on power system reliability evaluation. The investigation includes long-term system planning and short-term operational planning on reliability evaluation taking into accounts different penetration levels of wind power. This thesis presents a Wind-Hydro Cooperation (WHC) method for reducing the effects of wind power fluctuations on system operation and maintaining similar power system reliability level at different wind penetrations. The thesis starts by gaining insights into methodologies of power system reliability evaluation. These methodologies are described and discussed in details. Then, relations between the wind speed and the wind turbine power output are explored fo r modelling the wind farm output. The output fluctuation of wind power affects the power system operation. The system requires additional operating reserve to maintain the original system reliability. To mitigate these effects of wind output fluctuation, the study explores the use of Pump-Hydro technology to cooperate with wind power to meet the increased operating reserve requirements. In order to verify and analyze the validity of the WHC method on reducing the effect of wind power fluctuation on power output and maintaining the system reliability, an extensive set of case studies are performed. Following this, the reliability analyses focus on three aspects: a) a small system for initial validation of the idea; b) a larger system for analysing the effects of the WHC method on long-term system planning and short-term operational planning; c) application of a real practical system. All the above tests involve investigation of wind penetration at different level starting from 10% to 40%. The period of investigation ranges from one day, one week, one month and then extending it to one full year.

The novelty of this research lies in the development of a dry, ultra-thin gate dielectric based on an inorganic/organic bi-layer with a total thickness of up to ~ 20 nm. The inorganic layer is aluminium oxide formed by UV/ozone exposure of aluminium layers. The organic layer is 1-octylphosphonic acid prepared by vacuum evaporation. A fully-dry fabrication process has been developed and the low-voltage OTFT operation has been demonstrated. In addition, the preparation of the aluminium oxide has been optimized through its implementation into p-channel thin-film transistors based on thermally evaporated pentacene. Results demonstrate that the UV/ozone exposure time primarily affects the threshold voltage of the transistors and the bias-stress induced shift in the threshold voltage. Both parameters improve when longer UV/ozone exposure times are implemented. Except for the lower field-effect mobility, the resultant transistor parameters are comparable to similar transistor structures reported to date using mixed wet-and-dry processes.

Pulsed power flash radiography is a rapidly growing technology which involves the generation of intense and short bursts of electric pulses to generate high energy X-rays to inspect dense objects. The generation of fast transition, short width, repetitive pulses for effective output requires switching technology which has limitations. From the traditional slow switching gas trigger switches, switching technology has evolved to semiconductors which has led to improved switching characteristics and delivered pulse shapes. One such semiconductor which is of interest in flash radiography is a power MOSFET. Due to superior switching speeds, repetition rates and modularity, MOSFETs are a strong contender in future pulsed power applications. This thesis focuses on the design of a MOSFET based pulsed power system capable of generating 2.2kV, 200A pulses with turn on and turn off times of less than 10ns and 20ns respectively. In order to achieve the target specification, a high speed current source MOSFET gate driver design specific to the application is proposed. Further, an inductive voltage adder system which utilises multiple lower voltage pulse sources driven by MOSFETs and outputs a higher voltage which is the summation of the total number of voltage sources, is designed to be compatible with the proposed gate driver scheme. A compact high power density pulsed power prototype is demonstrated which provides a benchmark for future research into the development of a larger scale system for flash radiography.

Cognitive Wireless Sensor Networks (CogWSNs) are an adaptive learning based wireless sensor network relying on cognitive computational processes to provide a dynamic capability in configuring the network. The network is formed by sensor nodes equipped with cognitive modules with awareness of their operating environment. If the performance of the sensor network does not meet requirements during operation, a corrective action is derived from stored network knowledge to improve performance. After the action is implemented, feedback on the action taken is evaluated to determine the level of improvement. Example functions within CogWSNs can be as simple as to provide robust connectivity or as complex as to negotiate additional resources from neighbouring network groups with the goal of forwarding application-critical data. In this work, the concept of CogWSNs is defined and its decision processes and supporting architecture proposed. The decision role combines the Problem Solving cognitive process from A Layered Reference Model of the Brain and Polya Concept, consisting of Observe, Plan, Implement, and Evaluate phases. The architecture comprises a Transceiver, Transducer, and Power Supply virtual modules coordinated by the CogWSN's decision process together with intervention, if necessary, by a user. Three types of CogWSN modules are designed based on different implementation considerations: Rule-based CogWSN, Supervised CogWSN, and Reinforcement CogWSN. Verification and comparison for these modules are executed through case studies with focus on power transmission and communication slot allocation. Results show that all three modules are able to achieve targeted connectivity and maintain utilisation of slots at acceptable data rates.

A large variety of sources will provide energy, including nuclear fusion energy from tokamak type reactors; still at a research stage of development to produce 500 MW of energy. The reactor's structure will be affected by temperature changes and ionising radiation (neutrons produced by the burning plasma); it is therefore necessary to instrument the structure with long lasting, accurate, temperature and strain monitoring devices. This thesis demonstrates that using electrical sensors in the supporting structures of a fusion reactor for measuring strain and temperature will be strongly affected by the ionising radiation and the electromagnetic fields. This is because it was found that errors of over 100% can be present on the measurements due to combined electromagnetic and radiation effects. An optical fibre sensor was constructed to a strain resolution of ±10uε and temperature resolution of ±5°C. The fibre sensor is an off the shelf solution dual-wavelength fibre Bragg grating. The thesis proposes a method of coating the fibre in layers of silver and copper to protect this sensor. A new method of embedding the fibre in a stainless steel piece is described using a silver solder to couple the fibre with the steel increasing its temperature and strain sensitivity by approximately three times, with the solder melted using an induction heater. To measure the temperature and strain response from the fibre Bragg grating a broadband optical filter was used as a real-time interrogation system. Data capture and analysis software solution was created running at 30Hz and ~1pm resolution; temperature and strain requirements were met. This work describes a method of performing measurements of strain and temperature not subject to electromagnetic interference, with good electrical isolation possible and capable of operating in a near vacuum. The sensor package provides a robust, weldable solution for connecting the sensor to the port plug structure and/or plasma instrumentation equipment that will provide regular repeatable measurements of strain and temperature.

A viable last-mile solution for high-speed high-capacity optical networks capable of securely supporting a large number of simultaneous users by minimal hardware requirements is needed. Optical code division multiple access (OCDMA) is an advanced multiplexing scheme which provides a more efficient and fairer division of available bandwidth among users in comparison to other techniques such as wavelength division multiplexing (WDM) and optical time division multiplexing (OTDM). As a result, OCDMA has become a very attractive multi-access technique that can be used in local area networks (LAN) and last-mile application. I have built an incoherent OCDMA system, whose performance I investigated, including the influence of chromatic dispersion, timing jitter and the effect of ambient temperature fluctuations on data transmission over a 17-km bidirectional fibre link (a testbed) I designed between Strathclyde and Glasgow University. The OCDMA system is based on two-dimensional wavelength-hopping time-spreading (2D-WH/TS) incoherent OCDMA codes which use multiwavelength picosecond pulses. The encoders/decoders are based on fibre Bragg gratings (FBG) technology and the investigations were carried out at OC-48 (2.5 Gb/s) data rate. The testbed was built to be a fully chromatic dispersion compensated with sub-picosecond accuracy. Synchronisation and timing jitter effects were investigated on OCDMA signal transmitted over 17-km distance. A receiver that incorporates an all-optical clock recovery (AOCR) for synchronisation in incoherent OCDMA systems was developed and demonstrated. The all-optically recovered clock signal was then used to drive optical "time gate" to control a switching window ("a time gate"). This is to pass the autocorrelation peak while blocking the MAI noise to improve system power budget and overall performance.

An externalized sound is one that is perceived outside the head, whereas an internalized sound is perceived inside the head. The effect of hearing aids on these phenomena was investigated through psychoacoustic experiments, a novel questionnaire and offline analyses. The importance of high-frequency pinna cues for externalization in normal-hearing and hearing-impaired listeners was investigated using open-ear simulations of different microphone placements, frequency responses, and the number of talkers. It was found that hearing-impaired listeners experienced a compressed or "flattened" perception of externalization in relation to pinna cues. The role of changes in the stimulus and source direction on the perception of externalization by hearing-impaired listeners with and without their hearing aids was also investigated. An effect of angle but no effect of hearing aids was found for hearing-impaired listeners. The effect of short-term acclimatization to hearing aids by normal-hearing listeners performing the same task was investigated; an effect of acclimatization was found. A questionnaire was developed to determine the prevalence of internalization in several situations. The prevalence of internalization increased with the number of hearing aids worn. The overall prevalence for any experience of internalization was 25% of the sample population. The effect of dynamic-range compression, signal type and listening environment on a potential indicator for internalization, the shape of the interaural-level-difference distribution, was analyzed. The analyses only indicated potential internalization under particular constraints related to listening environment and temporal resolution, not dynamic range compression.