Spintronics is a new and emerging field in nanotechnology, which has been evolving rapidly. It aims to exploit the spin degree of freedom in order to realise advanced electronic devices. With the recent improvement in the storage media devices following to the discovery of giant magnetoresistance effect, it is envisioned that the electronic devices have several advantages over the conventional electronics in respect to the storage capacity, speed and power consumption. One avenue towards next generation spintronic devices is to develop half-metallic ferromagnets (HMFs) with 100% spin polarisation (P) and Curie temperature (TC) above room temperature (RT). HMFs have unique properties, in which the majority spins have a metallic band structure, whereas the minority spins have a semiconducting band with the Fermi level (EF) lying within an energy gap. P of HMFs has been predominantly estimated using Julliere’s formula in a magnetic tunnel junction (MTJ) or measured by the Andreev reflection (AR) at low temperature. Both methods are very sensitive to the surface/interface spin polarisation. Alternative optical methods such as photoemission have also been employed. However, these methods require a complicated and expensive set-up. Therefore, it is of paramount importance to directly and easily measure the band-gap of HMFs. The main aim of this study is to develop a new technique to directly measure the band-gap (Eg) of HMFs at RT. For that, a simple experimental set-up has been designed utilising circularly-polarised infrared (IR) light to excite minority spins and measure the corresponding absorption. The validity of our technique was demonstrated by measuring Eg of polycrystalline and epitaxial full-Heusler-alloy films and single-crystal half-Heusler alloy bulk. Eg measurements have revealed relevant figures to the calculated and expected reported values. This technique allows measuring Eg of HMFs at RT for the first time. It can therefore be used to provide simple optimisation of growth conditions.

At present, the battery is employed as a power source for wide varieties of microelectronic systems ranging from biomedical implants and sensor net-works to portable devices. However, the battery has several limitations and incurs many challenges for the majority of these systems. For instance, the design considerations of implantable devices concern about the battery from two aspects, the toxic materials it contains and its lifetime since replacing the battery means a surgical operation. Another challenge appears in wire-less sensor networks, where hundreds or thousands of nodes are scattered around the monitored environment and the battery of each node should be maintained and replaced regularly, nonetheless, the batteries in these nodes do not all run out at the same time. Since the introduction of portable systems, the area of low power designs has witnessed extensive research, driven by the industrial needs, towards the aim of extending the lives of batteries. Coincidentally, the continuing innovations in the field of micro-generators made their outputs in the same range of several portable applications. This overlap creates a clear oppor-tunity to develop new generations of electronic systems that can be powered, or at least augmented, by energy harvesters. Such self-powered systems benefit applications where maintaining and replacing batteries are impossi-ble, inconvenient, costly, or hazardous, in addition to decreasing the adverse effects the battery has on the environment. The main goal of this research study is to investigate energy harvesting aware design techniques for computational logic in order to enable the capa- II bility of working under non-deterministic energy sources. As a case study, the research concentrates on a vital part of all computational loads, SRAM, which occupies more than 90% of the chip area according to the ITRS re-ports. Essentially, this research conducted experiments to find out the design met-ric of an SRAM that is the most vulnerable to unpredictable energy sources, which has been confirmed to be the timing. Accordingly, the study proposed a truly self-timed SRAM that is realized based on complete handshaking protocols in the 6T bit-cell regulated by a fully Speed Independent (SI) tim-ing circuitry. The study proved the functionality of the proposed design in real silicon. Finally, the project enhanced other performance metrics of the self-timed SRAM concentrating on the bit-line length and the minimum operational voltage by employing several additional design techniques.

There is a significant number of visually impaired individuals who suffer sensitivity loss to high-spatial frequencies, for whom current optical devices are limited in degree of visual aid and practical application, Digital image and video processing offers a variety of effective visual enhancement methods that can be utilised to obtain a practical embedded augmented vision device such as a head mounted display device, Common approaches to augmented vision extract an images high-spatial frequencies through digital image processing edge detection techniques, which are then overlaid on top of the original image to improve visual perception amongst the visually impaired, Augmented visual aid devices require highly user-customisable, real-time capable algorithms designed for subjective configuration per task, where current digital image processing visual aids offer very little user-configurable options, Firstly, the effectiveness of various digital image edge detection techniques through augmented image visual experiments with simulated low-vision subjects are investigated, A comparitive study of the 6416 DSP and Virtex-5 FPGA embedded platforms are evaluated for performance benchmarks of implemented edge detectors of various complexity. In addition to optimising the mathematically complex statistical edge detection algorithm for embedded implementation. A highly user-customisable real-time morphology edge enhanced augmented vision algorithm and FPGA realisation are presented, where the edge type, magnitude and edge thickness can be modified during real-time operation. A reconfigurable morphological architecture for real-time implementation on FPGA is developed, which obtains performance comparable to other approaches, in addition to obtaining a significant degree of reconfigurability not previously demonstrated in literature. A morphological abstraction framework is presented, where images are significantly abstracted to obtain efficient visual entropy through enhancing key edge component information, while simultaneously reducing other image information. The morphological abstraction framework is highly suited for FPGA implementation, produces visually comparable results to equivalent methods, while obtaining efficient computational complexity. A morphological edge preserving smoothing filter is presented, which utilises adaptive structuring element functions obtained from a counter-harmonic mean bilateral filter that asymptotically corresponds to morphological operations.

A fundamental research problem of cooperative half-duplex relaying network is to find its capacity as well as a practical approach in order to achieve the capacity. In this thesis, the state-of-the-art review shows that the achievable rates have been derived as well as a capacity upper bound using max-flow min-cut theorem for the degraded channel with multiple parallel relays. However, the capacity of the relay channel for general case remains unknown. In the literatures, there are several relaying strategies having been developed to push the performance towards the capacity. However, the existing strategies can achieve the desired performance only when the full channel information is available at the source, which causes a large amount of signalling overhead and channel feedback. Another problem for the half-duplex relaying network is that it suffers a loss of spectral efficiency due to the orthogonality requirement for the relay transmission. Half-duplex orthogonal relay often utilizes dedicated resources to help source-to-destination communications, which results in reduced spectral efficiency and under utilization of the allocated bandwidth. The main contributions of this thesis are three-fold: • First, we propose a novel relay selection algorithm based on mixed channel information. The proposed selection algorithm aims to reduce the signalling overhead of modulation-adaptive Decode-and-Forward (DF) relaying by exploiting statistical channel information. A novel semi-deterministic approach is proposed to perform joint rate-adaptation and best-relay selection. It is shown that the proposed algorithm can achieve a good trade-off between the spectral efficiency and signalling overhead . ., Second, to mitigate the error propagation effects of DF relaying, we propose a distributed turbo decoding algorithm exploiting the source-relay correlation with reduced complexity. By iteratively updating the decoding output of the turbo decoders, the proposed decoding algorithm can outperform the conventional selective DF relaying. o Third, the relationship between the spectral efficiency and bandwidth dedicated to the relay is theoretically established. It is shown that, with the practical model of users' channel usage, the spectral efficiency is maximized when the relay does not have any dedicated bandwidth. Moreover, this theoretical result is elaborated through extensive investigation of the DF relay adopting various MAC-layer protocols including round robin resource scheduling, best-user selection, incremental relaying employing type II hybrid automatic repeat request, as well as joint channel and power allocation. Our Monte Carlo simulations show that the DF-relay with coordinated spectrum sharing improves the spectral efficiency by at least 50% in comparison with the relay with dedicated bandwidth allocation.

This thesis focussed upon the localisation of people and specifically the effect that the human body had upon estimated positions produced by the tested localisation algorithms. These algorithms used the signal strength measured between two nodes to provide ranging information to allow a user's position to be estimated. However, the human body causes non-line of sight situations, termed as the body shadowing effect. This effect is especially important in this work as the node worn by the user took on a smartwatch style form factor worn on the wrist. This natural positioning meant that line of sight between two nodes could be easily blocked by the user's body. Body shadowing caused deviation from the true positions which were generally skewed from the true path depending upon the orientation of the user. These deviations are detailed in this thesis and a number' of approaches were taken to reduce them. Firstly was the implementation of a statistically based localisation algorithm, which incorporated a history of the user's estimated positions into its design. This reduced the erratic trajectories produced, and ensured momentary non-line of sight would have a lesser effect upon estimated trajectories. Secondly was the implementation of line of sight specific channel models which were employed depending upon the user's orientation. Thirdly, two diversity schemes were investigated, polarisation and spatial. Polarisation diversity took on the form of two receivers at each node built into the infrastructure arranged 90° to each other. Spatial diversity saw the user wear two nodes, one on each wrist. Finally an extension to the statistically based localisation algorithm was presented that allowed cooperation between multiple users to enhance their localisation accuracy.

High Q coaxial waveguide devices are found in superior performance filters for wireless applications, they are used with cavity perturbation techniques to measure complex permittivity, they underpin the operation of scanning microwave microscopes and are of ongoing importance in the design of space division multiple access beamforming antennas. The coaxial 'T' junction transition formed between the feed and the coaxial waveguide airspace is a key junction in all of these waveguide coaxial systems. The study of this transition was the starting point for this research. In previous literature, a Green's theorem has been used to characterise a coaxial 'T' junction. In this thesis the model has been extended to include higher order propagating modes, evanescent modes, partial and full height probes, and any general form of coaxial termination. The final model is capable of analysing the S-parameters of multiport coaxial devices, and the energy distribution of the modes within the coaxial airspace. Several different coaxial systems were constructed and experimentally investigated including a coaxial resonator, a non-standard coaxial waveguide unknown load, and a beamforming antenna. In all cases the dyadic Green's function model was demonstrated to model accurately the performance of the multi-mode coaxial system. The major advantage of this approach is its inherent closed form nature lending itself to a computationally efficient means of formulating Sparameters, synthesising field patterns within the waveguide airspace and for shaping beam forming antenna radiation patterns. The model has successfully been applied to the design and experimental validation of the algorithm for several overmoded coaxial applications including: coaxial load characterisation, overmoded characteristic impedance formulations, overmoded resonant cavity and SDMA beamforming antenna design. Further applications to material measurement, direction of arrival estimations and mutual compensation matrices for antenna beamformers are under review. In essence, a powerful generic tool for the design and prediction of the electromagnetic characteristics of single and multiple probe overmoded coaxial devices has arisen from this investigation.

The work presented in this thesis covers the design of various sampling/ processing schemes which allow the capabilities of a digital spectral analysis system to be extended to cope with specific spectral analysis applications where sampling rates are restricted. It is shown how the sampling/processing methods can be tailored to suit various types of signals so as to minimise the sampling rate. Uniform sampling of multiple narrowband signals is the first specialised sampling technique to be discussed. Here sampling rate reductions are devised by employing spectral band interlacing which is similar in principle to the classical case of sampling a single narrowband signal but on a much more complicated level. Periodic staggered sampling is also considered with application to the sampling of multiple bandlimited signals. It is shown that such a sampling technique can be constructed with a mean sampling rate equal to the occupied frequency space of the signal, yet be insensitive to the signals frequency distribution of energy from the point of view of reconstruction ambiguity. To cover the spectral analysis of random signals a general resume is given to both statistical and random sampling techniques in the estimation of the power spectral density of stationary random waveforms. Statistical sampling is put forward as an unconventional way of digital spectral analysis in an attempt to break the "engineering" approach of uniform time sampling. Random sampling is suggested as a way of reducing the sampling rate for spectral analysis of stochastic waveforms but problems with spectral noise levels were encountered. A new approach to autocorrelation estimation is then given in an attempt to combat the defficiencies of random sampling but to maintain sampling rates as low as possible. The sampling/processing techniques involved with this new method are again based on a staggered periodic sampling waveform but of a specially designed staggered pattern. The thesis finishes with a short description of the practical problem of measuring turbine blade vibrations using blade-tip displacement detection. Suggestions are given as to the use of two of the data processing methods which are based on periodic staggered sampling. It is shown possible to use these two processing techniques in parallel while operating on a common source of data samples and giving separate output spectral analyses on different constituent vibrations present on the turbine blading.

We consider a nematic liquid crystal device in which a bistable surface anchoring term produces two stable states, a Vertical State (i. e. all molecules are homeotropically aligned) and a Hybrid Aligned Nematic (HAN) State (i. e. the molecules are homeotropically aligned on one boundary and homogeneously aligned on the other). Our one-dimensional model determines the director profile throughout a nematic cell by minimizing its free energy. The free energy in this model contains dielectric, elastic, flexoelectric and anchoring terms. This constitutes what we denote the 'no-flow' model. An expanded, so-called 'flow model', also includes a flow equation that we couple with our system of director equations. We then introduce three time integration methods for our numerical simulations, namely an explicit method, a semi-implicit method and a fully-implicit method, each of which employs an adaptive time-stepping algorithm to control the size of each time-step. Numerical simulations also employ a moving mesh algorithm to control the positioning and quantity of node points used at each time-step. We then compare each simulation method to determine which provides the optimal balance of speed and accuracy. We investigate switching for voltage pulses of different magnitude and duration in order to graph standard rV-plots. Each switching region is determined by the interaction between the bistable surface and bulk equations once the applied voltage is removed, which is a relatively complex process. We develop and present a powerful algorithm for automatically generating rV-plots corresponding to any given parameter set. Using this algorithm, we then investigate the effect of each parameter on the switching characteristics of our cell, using both the standard model and the expanded 'flow' model. The effects of flow are investigated by comparing the results of each model via numerical simulation. We show that flow-induced kickback in the director can significantly affect the results obtained using a no-flow model.

No description available.

No description available.