Audio signals, such as speech and ambient sounds convey rich information pertaining to a user’s activity, mood or intent. Enabling machines to understand this contextual information is necessary to bridge the gap in human-machine interaction. This is challenging due to its subjective nature, hence, requiring sophisticated techniques. This dissertation presents a set of computational methods, that generalize well across different conditions, for speech-based applications involving emotion recognition and keyword detection, and ambient sounds-based applications such as lifelogging.

The expression and perception of emotions varies across speakers and cultures, thus, determining features and classification methods that generalize well to different conditions is strongly desired. A latent topic models-based method is proposed to learn supra-segmental features from low-level acoustic descriptors. The derived features outperform state-of-the-art approaches over multiple databases. Cross-corpus studies are conducted to determine the ability of these features to generalize well across different databases. The proposed method is also applied to derive features from facial expressions; a multi-modal fusion overcomes the deficiencies of a speech only approach and further improves the recognition performance.

Besides affecting the acoustic properties of speech, emotions have a strong influence over speech articulation kinematics. A learning approach, which constrains a classifier trained over acoustic descriptors, to also model articulatory data is proposed here. This method requires articulatory information only during the training stage, thus overcoming the challenges inherent to large-scale data collection, while simultaneously exploiting the correlations between articulation kinematics and acoustic descriptors to improve the accuracy of emotion recognition systems.

Identifying context from ambient sounds in a lifelogging scenario requires feature extraction, segmentation and annotation techniques capable of efficiently handling long duration audio recordings; a complete framework for such applications is presented. The performance is evaluated on real world data and accompanied by a prototypical Android-based user interface.

The proposed methods are also assessed in terms of computation and implementation complexity. Software and field programmable gate array based implementations are considered for emotion recognition, while virtual platforms are used to model the complexities of lifelogging. The derived metrics are used to determine the feasibility of these methods for applications requiring real-time capabilities and low power consumption.

This work describes the development of automated flows to generate pad rings, mixed signal power grids, and mega cells in a multi-project test chip. There were three major design flows that were created to create the test chip. The first was the pad ring which was used as the staring block for creating the test chip. This flow put all of the signals for the chip in the order that was wanted along the outside of the die along with creation of the power ring that is used to supply the chip with a robust power source.

The second flow that was created was used to put together a flash block that is based off of a XILIX XCFXXP. This flow was somewhat similar to how the pad ring flow worked except that optimizations and a clock tree was added into the flow. There was a couple of design redoes due to timing and orientation constraints.

Finally, the last flow that was created was the top level flow which is where all of the components are combined together to create a finished test chip ready for fabrication. The main components that were used were the finished flash block, HERMES, test structures, and a clock instance along with the pad ring flow for the creation of the pad ring and power ring.

Also discussed is some work that was done on a previous multi-project test chip. The work that was done was the creation of power gaters that were used like switches to turn the power on and off for some flash modules. To control the power gaters the functionality change of some pad drivers was done so that they output a higher voltage than what is seen in the core of the chip.

An important operating aspect of all transmission systems is power system stability

and satisfactory dynamic performance. The integration of renewable resources in general, and photovoltaic resources in particular into the grid has created new engineering issues. A particularly problematic operating scenario occurs when conventional generation is operated at a low level but photovoltaic solar generation is at a high level. Significant solar photovoltaic penetration as a renewable resource is becoming a reality in some electric power systems. In this thesis, special attention is given to photovoltaic generation in an actual electric power system: increased solar penetration has resulted in significant strides towards meeting renewable portfolio standards. The impact of solar generation integration on power system dynamics is studied and evaluated.

This thesis presents the impact of high solar penetration resulting in potentially

problematic low system damping operating conditions. This is the case because the power system damping provided by conventional generation may be insufficient due to reduced system inertia and change in power flow patterns affecting synchronizing and damping capability in the AC system. This typically occurs because conventional generators are rescheduled or shut down to allow for the increased solar production. This problematic case may occur at any time of the year but during the springtime months of March-May, when the system load is low and the ambient temperature is relatively low, there is the potential that over voltages may occur in the high voltage transmission system. Also, reduced damping in system response to disturbances may occur. An actual case study is considered in which real operating system data are used. Solutions to low damping cases are discussed and a solution based on the retuning of a conventional power system stabilizer is given in the thesis.

In this work a newly fabricated organic solar cell based on a composite of fullerene derivative [6,6]-phenyl-C61 butyric acid methyl ester (PCBM) and regioregular poly (3-hexylthiophene) (P3HT) with an added interfacial layer of AgOx in between the PEDOT:PSS layer and the ITO layer is investigated. Previous equivalent circuit models are discussed and an equivalent circuit model is proposed for the fabricated device. Incorporation of the AgOx interfacial layer shows an increase in fill factor (by 33%) and power conversion efficiency (by 28%). Moreover proper correlation has been achieved between the experimental and simulated I-V plots. The simulation shows that device characteristics can be explained with accuracy by the proposed model.

GaAs single-junction solar cells have been studied extensively in recent years, and have reached over 28 % efficiency. Further improvement requires an optically thick but physically thin absorber to provide both large short-circuit current and high open-circuit voltage. By detailed simulation, it is concluded that ultra-thin GaAs cells with hundreds of nanometers thickness and reflective back scattering can potentially offer efficiencies greater than 30 %. The 300 nm GaAs solar cell with AlInP/Au reflective back scattering is carefully designed and demonstrates an efficiency of 19.1 %. The device performance is analyzed using the semi-analytical model with Phong distribution implemented to account for non-Lambertian scattering. A Phong exponent m of ~12, a non-radiative lifetime of 130 ns, and a specific series resistivity of 1.2 Ω·cm2 are determined.

Thin-film CdTe solar cells have also attracted lots of attention due to the continuous improvements in their device performance. To address the issue of the lower efficiency record compared to detailed-balance limit, the single-crystalline Cd(Zn)Te/MgCdTe double heterostructures (DH) grown on InSb (100) substrates by molecular beam epitaxy (MBE) are carefully studied. The Cd0.9946Zn0.0054Te alloy lattice-matched to InSb has been demonstrated with a carrier lifetime of 0.34 µs observed in a 3 µm thick Cd0.9946Zn0.0054Te/MgCdTe DH sample. The substantial improvement of lifetime is due to the reduction in misfit dislocation density. The recombination lifetime and interface recombination velocity (IRV) of CdTe/MgxCd1-xTe DHs are investigated. The IRV is found to be dependent on both the MgCdTe barrier height and width due to the thermionic emission and tunneling processes. A record-long carrier lifetime of 2.7 µs and a record-low IRV of close to zero have been confirmed experimentally.

The MgCdTe/Si tandem solar cell is proposed to address the issue of high manufacturing costs and poor performance of thin-film solar cells. The MBE grown MgxCd1-xTe/MgyCd1-yTe DHs have demonstrated the required bandgap energy of 1.7 eV, a carrier lifetime of 11 ns, and an effective IRV of (1.869 ± 0.007) × 103 cm/s. The large IRV is attributed to thermionic-emission induced interface recombination. These understandings can be applied to fabricating the high-efficiency low-cost MgCdTe/Si tandem solar cell.

Detection of molecular interactions is critical for understanding many biological processes, for detecting disease biomarkers, and for screening drug candidates. Fluorescence-based approach can be problematic, especially when applied to the detection of small molecules. Various label-free techniques, such as surface plasmon resonance technique are sensitive to mass, making it extremely challenging to detect small molecules. In this thesis, novel detection methods for molecular interactions are described.

First, a simple detection paradigm based on reflectance interferometry is developed. This method is simple, low cost and can be easily applied for protein array detection.

Second, a label-free charge sensitive optical detection (CSOD) technique is developed for detecting of both large and small molecules. The technique is based on that most molecules relevant to biomedical research and applications are charged or partially charged. An optical fiber is dipped into the well of a microplate. It detects the surface charge of the fiber, which does not decrease with the size (mass) of the molecule, making it particularly attractive for studying small molecules.

Third, a method for mechanically amplification detection of molecular interactions (MADMI) is developed. It provides quantitative analysis of small molecules interaction with membrane proteins in intact cells. The interactions are monitored by detecting a mechanical deformation in the membrane induced by the molecular interactions. With this novel method small molecules and membrane proteins interaction in the intact cells can be detected. This new paradigm provides mechanical amplification of small interaction signals, allowing us to measure the binding kinetics of both large and small molecules with membrane proteins, and to analyze heterogeneous nature of the binding kinetics between different cells, and different regions of a single cell.

Last, by tracking the cell membrane edge deformation, binding caused downstream event – granule secretory has been measured. This method focuses on the plasma membrane change when granules fuse with the cell. The fusion of granules increases the plasma membrane area and thus the cell edge expands. The expansion is localized at the vesicle release location. Granule size was calculated based on measured edge expansion. The membrane deformation due to the granule release is real-time monitored by this method.

In mesoscopic physics, conductance fluctuations are a quantum interference phenomenon that comes from the phase interference of electron wave functions scattered by the impurity disorder. During the past few decades, conductance fluctuations have been studied in various materials including metals, semiconductors and graphene. Since the patterns of conductance fluctuations is related to the distributions and configurations of the impurity scatterers, each sample has its unique pattern of fluctuations, which is considered as a sample fingerprint. Thus, research on conductance fluctuations attracts attention worldwide for its importance in both fundamental physics and potential technical applications. Since early experimental measurements of conductance fluctuations showed that the amplitudes of the fluctuations are on order of a universal value (e2/h), theorists proposed the hypothesis of ergodicity, e.g. the amplitudes of the conductance fluctuations by varying impurity configurations is the same as that from varying the Fermi energy or varying the magnetic field. They also proposed the principle of universality; e.g., that the observed fluctuations would appear the same in all materials. Recently, transport experiments in graphene reveal a deviation of fluctuation amplitudes from those expected from ergodicity.

Thus, in my thesis work, I have carried out numerical research on the conductance fluctuations in GaAs nanowires and graphene nanoribbons in order to examine whether or not the theoretical principles of universality and ergodicity hold. Finite difference methods are employed to study the conductance fluctuations in GaAs nanowires, but an atomic basis tight-binding model is used in calculations of graphene nanoribbons. Both short-range disorder and long-range disorder are considered in the simulations of graphene. A stabilized recursive scattering matrix technique is used to calculate the conductance. In particular, the dependence of the observed fluctuations on the amplitude of the disorder has been investigated. Finally, the root-mean-square values of the amplitude of conductance fluctuations are calculated as a basis with which to draw the appropriate conclusions. The results for Fermi energy sweeps and magnetic field sweeps are compared and effects of magnetic fields on the conductance fluctuations of Fermi energy sweeps are discussed for both GaAs nanowires and graphene nanoribbons.

Learning by trial-and-error requires retrospective information that whether a past action resulted in a rewarded outcome. Previous outcome in turn may provide information to guide future behavioral adjustment. But the specific contribution of this information to learning a task and the neural representations during the trial-and-error learning process is not well understood. In this dissertation, such learning is analyzed by means of single unit neural recordings in the rats' motor agranular medial (AGm) and agranular lateral (AGl) while the rats learned to perform a directional choice task. Multichannel chronic recordings using implanted microelectrodes in the rat's brain were essential to this study. Also for fundamental scientific investigations in general and for some applications such as brain machine interface, the recorded neural waveforms need to be analyzed first to identify neural action potentials as basic computing units. Prior to analyzing and modeling the recorded neural signals, this dissertation proposes an advanced spike sorting system, the M-Sorter, to extract the action potentials from raw neural waveforms. The M-Sorter shows better or comparable performance compared with two other popular spike sorters under automatic mode. With the sorted action potentials in place, neuronal activity in the AGm and AGl areas in rats during learning of a directional choice task is examined. Systematic analyses suggest that rat's neural activity in AGm and AGl was modulated by previous trial outcomes during learning. Single unit based neural dynamics during task learning are described in detail in the dissertation. Furthermore, the differences in neural modulation between fast and slow learning rats were compared. The results show that the level of neural modulation of previous trial outcome is different in fast and slow learning rats which may in turn suggest an important role of previous trial outcome encoding in learning.

Animals learn to choose a proper action among alternatives according to the circumstance. Through trial-and-error, animals improve their odds by making correct association between their behavioral choices and external stimuli. While there has been an extensive literature on the theory of learning, it is still unclear how individual neurons and a neural network adapt as learning progresses. In this dissertation, single units in the medial and lateral agranular (AGm and AGl) cortices were recorded as rats learned a directional choice task. The task required the rat to make a left/right side lever press if a light cue appeared on the left/right side of the interface panel. Behavior analysis showed that rat's movement parameters during performance of directional choices became stereotyped very quickly (2-3 days) while learning to solve the directional choice problem took weeks to occur. The entire learning process was further broken down to 3 stages, each having similar number of recording sessions (days). Single unit based firing rate analysis revealed that 1) directional rate modulation was observed in both cortices; 2) the averaged mean rate between left and right trials in the neural ensemble each day did not change significantly among the three learning stages; 3) the rate difference between left and right trials of the ensemble did not change significantly either. Besides, for either left or right trials, the trial-to-trial firing variability of single neurons did not change significantly over the three stages. To explore the spatiotemporal neural pattern of the recorded ensemble, support vector machines (SVMs) were constructed each day to decode the direction of choice in single trials. Improved classification accuracy indicated enhanced discriminability between neural patterns of left and right choices as learning progressed. When using a restricted Boltzmann machine (RBM) model to extract features from neural activity patterns, results further supported the idea that neural firing patterns adapted during the three learning stages to facilitate the neural codes of directional choices. Put together, these findings suggest a spatiotemporal neural coding scheme in a rat AGl and AGm neural ensemble that may be responsible for and contributing to learning the directional choice task.

Nanolasers represents the research frontier in both the areas of photonics and nanotechnology for its interesting properties in low dimension physics, its appealing prospects in integrated photonics, and other on-chip applications. In this thesis, I present my research work on fabrication and characterization of a new type of nanolasers: metallic cavity nanolasers. The last ten years witnessed a dramatic paradigm shift from pure dielectric cavity to metallic cavity in the research of nanolasers. By using low loss metals such as silver, which is highly reflective at near infrared, light can be confined in an ultra small cavity or waveguide with sub-wavelength dimensions, thus enabling sub-wavelength cavity lasers. Based on this idea, I fabricated two different kinds of metallic cavity nanolasers with rectangular and circular geometries with InGaAs as the gain material and silver as the metallic shell. The lasing wavelength is around 1.55 μm, intended for optical communication applications. Continuous wave (CW) lasing at cryogenic temperature under current injection was achieved on devices with a deep sub-wavelength physical cavity volume smaller than 0.2 λ3. Improving device fabrication process is one of the main challenges in the development of metallic cavity nanolasers due to its ultra-small size. With improved fabrication process and device design, CW lasing at room temperature was demonstrated as well on a sub-wavelength rectangular device with a physical cavity volume of 0.67 λ3. Experiments verified that a small circular nanolasers supporting TE¬01 mode can generate an azimuthal polarized laser beam, providing a compact such source under electrical injection. Sources with such polarizations could have many special applications. Study of digital modulation of circular nanolasers showed that laser noise is an important factor that will affect the data rate of the nanolaser when used as the light source in optical interconnects. For future development, improving device fabrication processes is required to improve device performance. In addition, techniques need to be developed to realize nanolaser/Si waveguide integration. In essence, resolving these two critical issues will finally pave the way for these nanolasers to be used in various practical applications.