The protection of wireless communications links against interference is a key concern in mission critical systems. In particular low probability of interception (LPI) systems which operate below the thermal noise floor of the receiver are particularly vulnerable. To protect against interference it is often necessary to include some form of active interference suppression. Broadband adaptive beamforming is one such technique which may be used to suppress interference by spatial and spectral filtering of the signals received by an array antenna. The hardware requirements of adaptive Beamforming systems are high in comparison to other approaches. As a result, the total number of adaptable weights realisable in the beamformer may be limited by the size, weight and power constraints of the system. This will limit the degrees of freedom in the beamformer and hence, the interference cancellation capability of the beamformer. The effectiveness of increasing the number of time-taps in space-time adaptive processors (STAP) as a method of increasing the number of interferers the beamformer can simultaneously cancel in an environment containing a mix of narrowband, partialband and broadband interference is studied. An alternative scheme to free up degrees of freedom in the beamformer is proposed based on frequency-domain excision to pre-filter narrowband interference before it reaches the beamformer. This approach frees up degrees of freedom in the beamformer, which would otherwise be consumed by the narrowband interference, for use in cancelling partialband and broadband interference. To enable the excision filters to identify narrowband interference while allowing partialband and broadband interference to pass through, a novel scheme is presented which produces a frequency mask that varies on a per-bin basis. Eigenanalysis of the beamformer's covariance matrix is used to explore the ability of frequency-domain excision to desensitise the beamformer to narrowband interference while bit error rate (BER) simulations demonstrate the enhanced interference protection the scheme affords an LPI communications link. Mismatches between the gain and phase responses of the radio-frequency (RF) front-end channels in broadband adaptive beamformers can limit their cancellation performance against higher power partialband and broadband interference. This performance limitation arises from the decorrelation experienced by the received signals across the array due to the interchannel mismatches. In STAP systems this performance limitation may be mitigated by increasing the number of taps per channel in the STAP. However, the computational complexity of adding additional time-taps tends to be high in STAP beamformers. Two new methods of interchannel mismatch compensation are proposed based on efficient frequency-domain methods. Simulations using software models and experiments using a hardware STAP system demonstrate the proposed techniques' ability to improve the cancellation performance where interchannel mismatches are limiting cancellation performance.

This thesis reports the development and potential applications of several direct UV written devices. The material platform of these devices is a silica-on-silicon composite, fabricated using flame hydrolysis deposition. The device development within this thesis considers wet etching, micromachining and sputtering processes to enhance the physical and chemical actuation of direct UV written planar Bragg gratings. Physical actuation has been investigated through thermo-optic and strain-optic means, chemical actuation has been achieved through evanescent field exposure. Combining the chemical and physical actuators developed, a monolithic lab-on-a-chip based device has been fabricated. The demonstrated chip has the ability to monitor pressure, flow, temperature and refractive index of a measurand, flowing through a microfluidic channel. A novel multimode interference (MMI) based device is also presented, which uses an adaptive Bragg period to define the side walls of the device. The reported structure shows a potential reduction in excess loss, compared to alternative raster scanned device geometries for power splitting operations.

Performing a high level vision is usually based on features extracted at low and intermediate levels of the process of perception of a visual scene. Segmentation and matching are instrumental tasks in producing comparable features in applications such as medical imaging, mining and oil extraction, gaming consoles, face, ear and gait biometrics, and etc. The ultimate goal of this study is to develop a fully functional prior aided segmentation framework to extract deformable shapes over a sequence of frames. This thesis acknowledges the demand by these applications for a robust and flexible approach which is particularly designed to extract deformable timely shape sequences. It is also recognised that existing methods are either too general, and thus inaccurate, or too specific, thereby limited in usability. This thesis suggests a learning model for gait synthesis with the ability to extrapolate to novel data. It involves computing comparable features from multiple sources. We show that these features which we formulate as continuous functions can be modelled by linear PCA. This thesis also proposes a new fast and robust shape registration algorithm to match shapes from different sources in the proposed framework. This algorithm is based on linear orthogonal transformations and shape moments. The registration parameters are computed directly by analysing the signed distance functions of the shapes. This is in-line with the level sets based prior shape segmentation framework adopted here. The segmentation is performed in a balanced framework between the data in the given images on one hand and the prior induced by the shape model and the registration algorithm proposed here on the other hand. This configuration ensures more control for the shape force over the overall shape geometry. Thus, favouring shapes familiar to the learned knowledge.

In modern multi-objective design optimisation, an effective geometry engine is becoming an essential tool and its performance has a significant impact on the entire process. Building a parametric geometry requires difficult compromises between the conflicting goals of robustness and flexibility. The work presents a solution for improving the robustness of parametric geometry models by capturing and modelling relative engineering knowledge into a surrogate model, and deploying it automatically for the search of a more robust design alternative while keeping the original design intent. Design engineers are given the opportunity to choose from a list of optimised designs to balance the robustness of the geometry and the original design intent. The prototype system is firstly tested on a 2D intake design repair example and shows the potential to reduce the reliance on human design experts in the conceptual design phase and improve the stability of the optimisation cycle. It also helps speed up the design process by reducing the time and computational power that could be wasted on flawed geometries or frequent human interferences. A case-study of the proposed repair system based on the design and analysis of a three-dimensional parametric turbine blade model has been set up. An automatic analysis workflow is set up and the results are summarised for setting up a repair database based on surrogate training methods. Positive repair results have been achieved and an automatic repair cycle for the blade model is being set up and tested. The proposed physics and engineering knowledge based geometry repair system for robust parametric geometries proves an effective tool for ensuring automation robustness and design flexibility.

The prospect of emulating the impressive computational capacities of biological systems has led to much interest in the design of analog circuits, potentially implementable in VLSI CMOS technology, that are guided by biologically motivated models. However, system design inevitably encounters the contrary constraints of size(or complexity) and computational power (or performance). From a high level design point of view, we believe that theoretical analysis of the model properties will undoubtedly benefit the implementation at a lower level. This thesis focuses on this simple aim to provide an extensive study of task-specific models based on dynamical systems in order to reduce model complexity or enhance the performance of algorithms. In many examples, it is the self-evolving dynamics of the model that allows the intrinsic parallel computations of algorithms, which are traditionally expressed by differential equations and systems. For instance, simple image processing tasks, such as the detection of edges in binary and grayscale images, have been performed by a reaction diffusion equation using the FitzHugh-Nagumo model as the reaction term in the previous work done by Kurata et al. (2008); Nomura et al. (2003, 2008, 2011b,a). Once the initial condition is correctly assigned according to a processed image, system states of this model will automatically evolve to the final result. From an application of view, the spatial distribution of system state can be regarded as a grid network with a proper discrete pattern; each network node becomes a FitzHugh-Nagumo type of neuron, while the diffusion term turns out to be the nearest couplings among them, where the coupling strength k is proportional to the original coefficient of diffusion D. So, one neuron (node) in the network deals with one pixel in the processed image. However, in previous study, this one-to-one mapping of image pixels to component neurons makes the size of the network a critical factor in any such implementation. The wrong edges are found due to the intrinsic mechanism of the algorithm when the diffused the processed image are used to pick out edges among the grayscale intensity levels and their most successful method solves this problem by a doubling of the size of the network. In the thesis, we propose two main improvements of the original algorithm in order for the smaller complexity and the better performance. We treat dynamics of the coupled system for the purpose of edge detection as a k-perturbation of the uncoupled one. Based on stability analysis of system state for both uncoupled and coupled cases, the system used for edge detection is identified as a Multiple Attractor type network and the final edge result corresponds to an attractor in high dimensional space. Hence, we conclude that the edge detection problem maps an image to an initial condition that is correctly located within the attraction domain of an expected attractor. For the first improvement, in order to get rid of the wrong edges, we provide a way of quantify the excitability of uncoupled neurons based on the Lyapunov exponents so that the boundary of attraction domain of the attractors can be well estimated. Moreover, an anisotropic diffused version of processed image is used for the further enhancement on the performance. For the second improvement, in order for diffusion of the processed image being accomplished by the hardware, we introduce a self-stopping mechanism to the original equation. Moreover, we link the basic design rules on system parameter settings to the fundamental theorem of WCNN (weakly coupled neural network) (Hoppensteadt and Izhikevich, 1997), which states that an uncoupled neuron must be near a threshold (bifurcation point) in order for rich dynamics to be presented in the coupled network. We apply our techniques to detect edges in data sets of artificially generated images (both noise-free and noise polluted) and real images, demonstrating performance that is as good if not better that the results of (Nomura et al., 2011b,a) with a smaller size of the network.

In this thesis, novel Foutain codes are proposed for transmission over wireless channels. The thesis concentrates on a specific version of Fountain codes, namely on Luby transform codes. More specifically, we consider their concatenation with classic error correction codes, yielding schemes, such as the concatenated Luby Transform and Bit Interleaved Coded Modulation using Iterative Decoding (LT-BICM-ID), the amalgamated Luby Transform and Generalized Low Density Parity Check (LT-GLDPC) code, or the Luby Transform coded Spatial Division Multiple Access (LT-SDMA) scenario considered. The thesis also investigates the potential of Systematic Luby Transform (SLT) codes using soft-bit decoding and analyses their Bit Error Ratio (BER) performance using EXtrinsic Information Transfer (EXIT) charts. SLT codes using different degree distributions and random integer packet index generation algorithms for creating the parity and information part of the SLT codeword are also investigated in this thesis. For the sake of improving both the BER performance and the diversity gain of Vertical Bell Laboratories Layered Space Time (V-BLAST) schemes, in this thesis a SLT coded V-BLAST system having four transmit and four receive antennas is proposed. Finally, A Hybrid Automatic Repeat reQuest (H-ARQ) SLT coded modulation scheme is designed in this thesis, where SLT codes are used both for correcting erroneous bits and for detecting as well as retransmitting erroneous Internet Protocol (IP) based packets. Erroneous IP packet detection is implemented using syndrome checking with the aid of the SLT codes’ Parity Check Matrix (PCM). Optimizing the mapping of SLT-encoded bits to modulated symbols and then using iterative decoding for exchanging extrinsic information between the SLT decoder and the demapper substantially improves the achievable Bit Error Ratio (BER) performance of the scheme.

A range of amorphous silicon based films and solar cells have been fabricated using an inductively coupled PECVD reactor in order to investigate how the deposition conditions influenced their properties. Complete cells were deposited onto rigid and flexible substrates, and homogeneous thin films were grown to investigate their optical and electrical properties separately from the influence of whole devices. Initial photoconductivity measurements performed on intrinsic amorphous silicon showed that it was suitable for photovoltaic applications (10-4 -1cm-1), complete single junction pin superstrate cells were then fabricated. The deposition conditions were altered to optimise cell efficiency, and the impact of substrate temperature, RF power, gas flow rate, system pressure, and substrate choice was commented on. The best cell produced was grown at 250C and was 8.05% efficient, after reversal of its structure to nip an efficiency of 5.77% was observed. The deposition temperature was lowered to 200C, and the properties of doped and intrinsic amorphous silicon films were investigated using measurements of the dark conductivity variation with temperature and spectroscopic ellipsometry. Optimum growth conditions were determined for each type of film and an nip cell was deposited using them resulting in an efficiency of 4.59%. The requirement of a carefully selected intrinsic region thickness in order to achieve effective collection of generated carriers was considered. Lastly the process was transfered from rigid glass to flexible polished stainless steel and Kapton substrates, requiring inversion of the cell orientation. A 1cm2 cell on a steel substrate was found to be functional under illumination and was 0.27% efficient. A 1cm2 cell was deposited onto Kapton but was only measured under dark conditions. It was concluded that the rear contact topography and defect density is of significant importance when dealing with flexible substrates and that sputtering rather than evaporation would have been a more suitable technique to fabricate it.

Many approaches to learning and teaching rely upon students working in groups. For- mation of optimal groups can be a time consuming and complex task, particularly when the list of participants is unknown in advance. This research investigates the imple- mentation of semantics to enhance computer-supported group formation in education using two approaches: The first approach uses semantics to express the criteria specified by the person forming the groups. The group formation in this approach is modelled as a constraint satisfaction problem where the criteria is a set of constraints that we aim to minimise their violation while processing the groups. The second approach uses Semantic Web domain ontologies in describing the participants to enrich the data used in calculating the similarity between the participants when the group formation is pro- cessed using a heuristic approach such as clustering algorithms. We run a number of experiments that include real datasets from higher education classes, simulated datasets, Web-based datasets, and user studies, to evaluate the re- search. The results proved that in both approaches, implementing semantics improved the generated groups, in that, using semantics to model group formation’s constraints generates an optimised grouping in terms of constraint satisfaction that exceeds the performance of existing applications, particularly in terms of the number of constraints it can handle; and that using semantics to model the participants’ data enhances their satisfaction with the groups they are allocated to.

Micro-Electro-Mechanical Systems (MEMS) design requires an integration of elements from two or more disparate physical domains: mechanical (translational, rotational, hydraulic), electrical, magnetic, thermal, etc. Different parts of a MEMS system are traditionally designed separately, using different methodologies and different tools applied to different energy domains. Although major Hardware Description Languages (HDLs) such as VHDL, Verilog and SystemC have been supplemented with analogue and mixed-signal (AMS) extensions which are essential in analogue and mixed-technology design, development of corresponding analogue and mixed-technology synthesis methodologies is still lagging behind. Therefore, there is an increasing need for automated synthesis techniques that can reduce the development cycle and facilitate the generation of optimal configurations. This research investigates and develops techniques for automated high-level performance optimisation and synthesis of mixed-technology MEMS systems. Results of this research have been published in 9 papers at peer reviewed international conferences and one two-part journal paper. Specific contributions of this research can be summarised as follows. Firstly, a dedicated distributed model of a mixed-technology MEMS case study of an accelerometer operating in a Sigma-Delta force-feedback control scheme is developed. The distributed behaviour is essential in the MEMS accelerometer design because it has been observed that sense finger resonance, usually not included in conventional models, affects the performance of the electromechanical Sigma-Delta feedback control. As shown in the simulation results, the Sigma-Delta loop failure, when the sense fingers bend seriously or oscillate, is captured by the proposed model but cannot be correctly modelled using conventional approach. Secondly, a novel, holistic approach is proposed for automated optimal layout synthesis of MEMS systems embedded in electronic control circuitry from user-defined high-level performance specifications and design constraints. The synthesis technique has been implemented in SystemC-A and named SystemC-AGNES. The method efficiently, and in an automated manner, generates suitable layouts of mechanical sensing element and configurations of the Sigma-Delta control loop by combining primitive components stored in a library and optimising them according to user specifications. Synthesis results show that the proposed technique explores the configuration space effectively, and it develops new structures which have not been investigated before. This contribution has been published as a two part paper in the Sensors & Transducers Journal. Finally, to enhance the modelling efficiency and capability of SystemC-A, for mixed-technology systems with crucial distributed behaviour, language extension has been proposed to efficiently support general partial differential equations(PDEs) modelling.

As a benefit of achieving a diversity gain and/or a multiplexing gain, MIMO techniques are capable of significantly increasing the achievable throughput and/or the network coverage without additional bandwidth or transmit power. For the sake of striking an attractive trade-off between the attainable diversity gain and/or multiplexing gain, in this thesis the novel Space-Time-Frequency Shift Keying (STFSK) concept is proposed for the family of MIMO systems. More specifically, in order to generate space-time-frequency domain codewords, the STFSK encoding schemes activate one out of Q dispersion matrices, and the associated address bits are then combined with a classic time-domain and frequency-domain modulation scheme. The resultant arrangements impose no inter-symbol interference and are capable of eliminating the inter-antenna interference, hence offering a range of benefits over other classic MIMO arrangements. Additionally, a soft-output STFSK demodulator is designed for iterative detection and the complexity of both the hard- as well as soft-decision demodulators is quantified. Furthermore, the STFSK performance is studied in both the single-user and the multipleuser multi-cell environment in order to investigate the effects of these techniques on the performance of the holistically optimized systems. Furthermore, we studied the H-ARQ systems advocated in the context of cooperation-aided wireless networks, where the MIMO elements are constituted by the individual elements of separate network nodes. Both perfect and imperfect coherent detection as well as non-coherent detection aided cooperative H-ARQ schemes are considered. In the perfect coherent detection based pilot symbol assisted scheme, a novel relay-switching aided H-ARQ scheme is proposed for mitigating the effects of correlation in fading wireless channels, followed by a H-ARQ scheme employing systematic Luby transform codes. In contrast to the unrealistic perfect coherent detection, realistic imperfect coherent schemes are studied, where the channel impulse responses are imperfectly estimated. Furthermore, non-coherent differential detection aided cooperative H-ARQ schemes are proposed and compared to their coherent detection assisted counterparts. Finally, a novel cooperative H-ARQ arrangement based on distributed space-time codes is proposed for the sake of improving the attainable system throughput, while reducing the system’s complexity.