In traditional radio frequency (RF) front-end subsystems, the passive components, such as antennas, filters, power dividers and duplexers, are separately designed and cascaded via the 50 Ω interfaces. This traditional approach results in a bulky and heavy RF front-end subsystem, and suffers from compromised efficiency due to the losses in the interconnections and the mismatching problems between different components. The frequency responses of the antennas such as the frequency selectivity and bandwidth are usually degraded, especially for microstrip antennas. To improve the frequency responses and reduce the size of RF front ends, it is important to investigate novel highly integrated antennas which exhibit multiple functions such as radiation, filtering, power dividing and combining or duplexing, simultaneously. In this thesis, several innovative designs of compact, multi-functional integrated an-tennas/arrays are proposed for wireless communication applications. First, new methods of designing integrated filtering antenna elements with broadband or dual-band performance are investigated. These antennas also feature high frequency selectivity and wideband harmonic suppression. Based on these studies, several integrated filtering array antennas with improved gains and frequency responses are developed for the first time. Compared with traditional array antennas, these proposed antennas exhibit improved bandwidths, out-of-band rejection and wideband harmonic suppression. The application of the filtering antennas in millimeter-wave (mm-Wave) frequency band is also investigated as it can potentially reduce the cost of the mm-Wave front-end subsystems significantly while providing the improved impedance bandwidth. The integrated design techniques are further developed to design novel dual-port highly integrated antennas with filtering and duplexing functions integrated. Such a new concept and the prototypes could find poten-tial applications in wireless communication systems and intelligent transportation system (ITS). In this thesis, comprehensive design methodologies and synthesis methods are provid-ed to guide the design of the integrated filtering antennas. The performance is evaluated with the help of full-wave electromagnetics (EM) simulations. All of the prototypes are fabricated and tested for validating the design concepts. Good agreement between the simulation and measurement results is achieved, demonstrating the integrated antennas have the advantages of compact size, flat gain performance, low losses and excellent harmonic suppression performance. These researches are important for modern wireless communication systems.

In this thesis, the algebras of primary interest are the quantum Schubert cells and the quantum Grassmannians, both of which are known to satisfy a condition on primitive ideals known as the Dixmier-Moeglin equivalence. A stronger version of the Dixmier-Moeglin equivalence is introduced - a version which deals with all prime ideals of an algebra rather than just the primitive ideals. Quantum Schubert cells are shown to satisfy the strong Dixmier-Moeglin equivalence. Until now, given a torus-invariant prime ideal of the quantum Grassmannian, one could not decide which quantum Plücker coordinates it contains. Presented here is a graph-theoretic method for answering this question. This may be useful for providing a full description of the inclusions between the torus-invariant prime ideals of the quantum Grassmannian and may lead to a proof that quantum Grassmannians satisfy the strong Dixmier-Moeglin equivalence.

Topological solitons are particle-like solutions of nonlinear field equations with important applications in physics. This thesis presents four research projects concerning topological solitons and their dynamics. We investigate solitons in (1+1)- and (2+1)-dimensions, and develop numerical methods to obtain static solutions and simulate soliton scattering. We first study kink collisions in a model with two scalar fields in the presence of false vacua. We find a variety of scattering outcomes depending on the initial velocity and vacuum structure. Kinks can either repel, form a true or false domain wall, annihilate, or collide and escape to infinity. These behaviours occur in alternating windows of initial velocity. When the kinks escape to infinity, there are a number of oscillations or 'bounces' before the kinks escape, and this bounce number is conserved in each of the windows. In the second project we design new baby Skyrme models that do not require a potential term to allow topological soliton solutions. We raise the Skyrme and sigma terms to fractional powers, which enables us to evade Derrick's theorem. We calculate topological energy bounds for our models and numerically obtain minimal energy solutions for solitons of charge 1, 2, and 3. For each charge, the minimal energy solution is a ring. The last two projects concern vortices in the Ginzburg-Landau model. In the first of these, we numerically investigate the scattering of multi-vortex rings. When two 2-vortex rings collide, there are two distinct scattering outcomes. In both cases, one pair of vortices will scatter at right angles and escape along the y-axis. The remaining two vortices will either form a bound state or escape along an axis after colliding a number of times. Finally, we study vortices scattering with magnetic impurities of the form σ(r)=ce-dr2 An impurity will attract or repel a vortex depending on the coupling constant λ and the parameters c and d. We scatter critically coupled vortices with two different impurities and explore the relationship between the scattering angle and impact parameter. We also find that a 2-vortex ring will break up in a head-on collision with an impurity.

Previous work has attempted to define the physiological determinants of load carriage over a single day of load carriage, primary aims of this thesis were to determine the physiological and biomechanical changes as a result of load carriage and to explore the causes of these changes during load carriage and to explore these changes over multiple days of activity. A secondary aim was to explore individual differences in performance and possible mechanisms for this. Chapter 4 observed acute changes in vertical ground reaction force, supported by effect sizes which suggests individual differences in performances (loading peak dCohens=1.66 and dGlass 4.49). These findings were supported by no change in first negative rate which suggests increased knee flexion is occurring to mitigate the effect of the load. Differences in anteroposterior ground reaction force variables suggest that changes in gait may affect movement economy. Energy expenditure was shown to be correlated to a number of strength variables, such as ankle plantarflexion (r=-0.47) and knee extensors(r=-0.46). Similar variables were studied as a result of 2 hours treadmill load carriage in chapter 5. An additional variable was the study of torque at specific joint angles in addition to peak torque. Drift was observed for V̇O2 (68.93%). The torque curves showed significant reduction for load carriage, around the optimum muscle lengths for force (Knee extension at 180°s-1: 95°-125°, knee flexion at 180°s-1: 95°-125°) with findings supported by the peak torque values, suggesting there is no shift in muscle function. These findings were supported by associations between knee extension neuromuscular function scores, ankle plantarflexion neuromuscular function scores and energy cost variables which suggest that reductions in neuromuscular function may account for the increased energy cost. Chapter 6 observed load carriage on repeated days. The chapter observed that energy cost did not recover to baseline 24 hours post day one (4.41%). Further increases in energy cost and V̇O2 were observed post day two for energy expenditure and V̇O2 which suggest that a cumulative increase in energy cost as a result of load carriage occurred. Knee extension at 60°s-1 and 0°s-1 was shown to not recover 24 hours post day one, these variables were shown to show increased reduction for post day two and day three. Similar findings were observed within the ankle plantarflexors but no significant changes were observed for knee flexion and ankle dorsiflexion. Variables were also studied in a field setting during a >12hour load carriage task by Greek Special Forces soldiers. This study observed increased Medial and lateral deviations of the centre of pressure which suggests ankle instability even during unloaded walking as a result of load carriage. Large reductions were observed in vertical jump height and power as a result of the task, however this did not correlate to any biomechanical findings.

Capture-recapture models have increased in complexity over the last decades and goodness-of-fit assessment is crucial to ensure that considered models provide an adequate fit to the data. In this thesis, my primary emphasis is to develop new diagnostic tools for capture-recapture models in order to target possible reasons of lack-of-fit, which might provide biological insights and point towards better-fitting candidate models. Starting with the basic Cormack-Jolly-Seber model, I develop a new tool for detecting heterogeneity in capture. I then progress to the more complex multi-state models, for which I propose a test for detecting a mover-stayer structure within the population. Finally, I move on to more general models presenting additional levels of uncertainty: first partial observations and then unobservable states. In the presence of partial observations, part of the observations are assigned to states with certainty whereas others are not. I develop a new test for the underlying state-structure of the partial observations, this test detects that the partial observations are not generated by the observable states defined in the experiment. In the presence of unobservable states, the additional level of uncertainty relates only to the non-captures. I present a procedure to test whether one or two unobservable states need to be defined in order for the model to provide an adequate fit to the data. Lastly, I explore the use of multi-state models to incorporate individual time-varying covariates, based on a fine discretisation of the covariate space.

Regenerative medicine is a rapidly expanding field of science with an exhaustive volume of literature published on the different strategies used to repair diseased or injured tissue. Recently, stem cells have emerged as a promising candidate in this regard owing to their involvement in embryogenesis, homeostatic turnover and normal tissue repair. Despite this potential, stem cell-based therapies have yet to be fully established in a clinical setting owing to complications associated with their limited numbers, immunogenicity, tumour formation and the ethical considerations surrounding their usage. Furthermore, the mechanisms underlying stem cell differentiation are complex and not fully understood, thus expanding stem cell numbers and predictably directing their commitment toward a desired lineage, represent a major challenge for tissue regeneration strategies. In an attempt to circumvent these problems there is currently a rising interest in biomimetic materials that aim to reproduce the physical architecture, chemical composition and plasticity of the in vivo extracellular environment in an in vitro setting. Furthermore, the need to expand stem cells while maintaining the stem cell phenotype has prompted many to look to the stem cell niche for answers. At the centre of most cellular responses to the physical cues embedded within the ECM are integrins. Integrins are mechanosensitive membrane spanning receptors that link the ECM to the cytoskeleton and thus transmit information from outside the cell into the nucleus, affecting gene transcription via a series of intracellular signalling cascades. To that end, many biomimetic systems incorporate integrin- binding ligands such as the tripeptide RGD. In this work glass surfaces functionalised with RGD were used to study changes in mesenchymal stem cell (MSC) responses to increased integrin binding by using an enzymatic ‘switch’ to reveal surface-bound RGD peptides that have been masked by a large chemical cap (Fmoc). The results of this work demonstrated that RGD- functionalised substrates can support MSC growth and influence them to commit to a particular fate. MSCs on surfaces where integrin-ligand binding was blocked developed a fibroblast-like phenotype whereas MSC grown on surfaces that were later enzymatically digested to reveal the underlying RGD ligands developed an osteoblast phenotype similar to RGD controls.

In this thesis light harvesting complexes, the LH2 and core complexes, from several different species of purple photosynthetic bacteria have been analysed both functionally and structurally. Purified monomeric core complexes from Rhodopseudomonas (Rps.) palustris have been used to isolate and identify the putative Protein W. This information was then used to create a Protein W deletion mutant. A low-resolution crystal structure of the monomeric core complex from Allochromatium (Alc.) vinosum is presented which suggests that the LH1 complex is a complete ellipse, unlike the core complex from Rps. palustris. It has previously been shown that some species are able to synthesise LH2 complexes that have different NIR absorption spectra. For example, strains 7050 and 7750 of Rps. acidophila can express both the B800-850 and B800-820 LH2 complexes, whilst strain 10050 only expresses the B800-850 LH2 complex despite evidence to suggest that this strain contains multiple LH2 genes (pucBA genes). It is this homogeneity that has made the LH2 complexes from this strain structurally amenable. Here, genomic DNA from Rps. acidophila strain 10050 has been isolated and sequenced using the next generation sequencing (NGS) technique, Illumina sequencing. So far 8 pucBA gene pairs were identified arranged into 2 distinct operons, one containing B800-850 pucBA genes and pucC, the putative Bchl transporter that is essential for efficient LH2 expression. The second operon contains B800-820 pucBA gene pairs only. Analysis of the protein products of the B800-850 type pucBA gene pairs has shown that none of these proteins match the sequence for the LH2 that is expressed by Rps. acidophila strain 10050. The crystal structure of the LH2 complex from the culture of Rps. acidophila used to isolate the genomic DNA was resolved to 2.05 Å from crystals of the LH2 complex. This structure shows that the protein sequence of the LH2 complex has not changed. Hence, not all the pucBA gene pairs have been identified in the genome sequence data. Currently mate-pair sequencing is being completed to fill in the gaps of sequence data and to complete the genome sequence. LH2 complexes contain carotenoid (Car) and Bchl molecules. In this thesis, the energy transfer mechanisms between Car and Bchl molecules have been investigated using 2-dimensional electronic spectroscopy (2DES). This technique splits the emission and excitation events on 2-dimensions, which can make the less populated ‘dark’ states more visible as overlapping peaks can be separated. Car moleucles are not seen as theoretically efficient in photosynthesis. This is due to short life times of the excited state S2. However, the Car used in photosynthesis have conjugated carbon tails with ≥9 π electrons. According to calculations by Tavan and Schulten, these molecules have the propensity to contain additional excited states that lie below the S2 state that can be involved in energy transfer and increase the efficiency of energy transfer between the Car and Bchl molecules. For the first time an intermediate Car electronic state has been directly observed and shown to be involved in energy transfer between the Car and Bchl molecules.

There is presently increased interest in using germanium (Ge) for both electronic and optical devices on top of silicon (Si) substrates to expand the functionality of Si technology. It has been extremely difficult to form an Ohmic contact to n-Ge due to Fermi level pinning just above the Valence band. A low temperature nickel process has been developed that produces Ohmic contacts to n-Ge with a specific contact resistivity of , which to date is a record. The low contact resistivity is attributed to the low resistivity NiGe phase, which was identified using electron diffraction in a transmission electron microscope. Light emission from Ge light emitting diodes (LEDs) was investigated. Ge is an indirect bandgap semiconductor but the difference in energy between the direct and indirect is small (~136 meV), through a combination of n-type doping and tensile strain, the band structure can be engineered to produce a more direct bandgap material. A silicon nitride (Si3N4) process has been developed that imparts tensile strain into the Ge. The stress in the Si3N4 film can be controlled by the RF power used during the plasma enhanced chemical vapour deposition. LEDs covered with Si3N4 stressors were characterised by Fourier transform infrared spectroscopy. Electroluminescence characterisation (EL) revealed that the peak position of the direct and indirect radiative transitions did not vary with the Si3N4 stressors due to the device geometries being too large. Therefore, nanostructures consisting of pillars smaller than a micron were investigated. Photoluminescence characterisation of 100 nm Ge pillars with Si3N4 stressors show emission at much longer wavelengths compared to bulk Ge (> 2.2 μm). In addition, the EL from Ge quantum wells grown on Si was also investigated. EL characterisation demonstrates two peaks around 1.55 and 1.8 μm, which corresponds to the radiative recombination between the direct and indirect transitions, respectively. This result is the first demonstration of EL above 1.45 μm for Ge quantum wells. Finally, the fabrication of Ge-on-Si single-photon avalanche detectors are presented. A single-photon detection efficiency of 4 % at 1310 nm wavelength was measured at low temperature (100 K). The devices have the lowest reported noise equivalent power for a Ge-on-Si single-photon avalanche detector (1×10-14 WHz-1/2).

We surveyed a number of genome visualisation tools used in biomedical research. We recognised that none of the tools shows all the relevant data geneticists who look for candidate disease genes would like to see. The biological researchers we collaborate with would like to view integrated data from a variety of sources and be able to see both data overviews and details. In response to this need, we developed a new visualisation tool, VisGenome, which allows the users to add their own data or data downloaded from other sources, such as Ensembl. VisGenome visualises single and comparative representations of the rat, the mouse, and the human chromosomes, and can easily be used for other genomes. In the context of VisGenome development we made the following research contributions. We developed a new algorithm (CartoonPlus) which allows the users to see different kinds of data in cartoon scaling depending on a selected basis. Also, two user studies were conducted: an initial quantitative user study and a mixed paradigm user study. The first study showed that neither Ensembl nor VisGenome fulfil all user requirements and can be regarded as user-friendly, as the users make a significant number of mistakes during data navigation. To help users navigate their data easily, we improved existing visualisation techniques in VisGenome and added a new technique CartoonPlus. To verify if this solution was useful, we conducted a second user study. We saw that the users became more familiar with the tool, and found new ways to use the application on its own and in connection with other tools. They frequently used CartoonPlus, which allowed them to see small regions of their data in a way that was not possible before.

Melanocytes within benign human nevi are the paradigm for tumor suppressive senescent cells in a pre-malignant neoplasm. These cells typically contain mutations in either the BRAF or NRAS oncogene and express markers of senescence, including p16INK4A 1-4. However, a nevus can contain 10s to 100s of thousands of clonal melanocytes and approximately 25% of melanomas are thought to arise in association with a pre-existing nevus 5-9. Neither observation is indicative of fail-safe senescence-associated proliferation arrest and tumor suppression. I set out to better understand the status of nevus melanocytes. Proliferation-promoting Wnt target genes, such as Cyclin D1 and c-Myc, were repressed in oncogene-induced senescent melanocytes in vitro, and repression of Wnt signaling in these cells induced a senescent-like state. In contrast, Cyclin D1 and c-Myc were expressed in many melanocytes of human benign nevi. Specifically, activated Wnt signaling in nevi correlated inversely with nevus maturation, an established dermatopathological parameter linked to clinical benignancy 10,11. Single cell analyses of lone interfollicular epidermal melanocytes and nevus melanocytes in tissue showed that expression of proliferation-promoting Wnt targets correlates with prior proliferative expansion of p16INK4A-expressing nevus melanocytes. In a mouse model, activation of Wnt signaling delayed, but did not bypass, senescence of oncogene-expressing melanocytes, leading to massive accumulation of proliferation-arrested, p16INK4A-positive, non-malignant melanocytes. I conclude that clonal hyperproliferation of oncogene-expressing melanocytes to form a nevus is facilitated by transient delay of senescence due to activated Wnt signaling. The observation that activation of Wnt signaling correlates inversely with nevus maturation, an indicator of lower malignant potential, supports the notion that persistent destabilization of senescence by Wnt signaling contributes to the malignant potential of nevi.