Engineering Commons^™

Bio-Inspired Learning And Hardware Acceleration With Emerging Memories, Shruti R. Kulkarni

Dissertations

Machine Learning has permeated many aspects of engineering, ranging from the Internet of Things (IoT) applications to big data analytics. While computing resources available to implement these algorithms have become more powerful, both in terms of the complexity of problems that can be solved and the overall computing speed, the huge energy costs involved remains a significant challenge. The human brain, which has evolved over millions of years, is widely accepted as the most efficient control and cognitive processing platform. Neuro-biological studies have established that information processing in the human brain relies on impulse like signals emitted by neurons called …

Analyzing Evolution Of Rare Events Through Social Media Data, Xiaoyu Lu

Dissertations

Recently, some researchers have attempted to find a relationship between the evolution of rare events and temporal-spatial patterns of social media activities. Their studies verify that the relationship exists in both time and spatial domains. However, few of those studies can accurately deduce a time point when social media activities are most highly affected by a rare event because producing an accurate temporal pattern of social media during the evolution of a rare event is very difficult. This work expands the current studies along three directions. Firstly, we focus on the intensity of information volume and propose an innovative clustering …

Blind Separation For Intermittent Sources Via Sparse Dictionary Learning, Annan Dong

Dissertations

Radio frequency sources are observed at a fusion center via sensor measurements made over slow flat-fading channels. The number of sources may be larger than the number of sensors, but their activity is sparse and intermittent with bursty transmission patterns. To account for this, sources are modeled as hidden Markov models with known or unknown parameters. The problem of blind source estimation in the absence of channel state information is tackled via a novel algorithm, consisting of a dictionary learning (DL) stage and a per-source stochastic filtering (PSF) stage. The two stages work in tandem, with the latter operating on …

Probabilistic Spiking Neural Networks : Supervised, Unsupervised And Adversarial Trainings, Alireza Bagheri

Dissertations

Spiking Neural Networks (SNNs), or third-generation neural networks, are networks of computation units, called neurons, in which each neuron with internal analogue dynamics receives as input and produces as output spiking, that is, binary sparse, signals. In contrast, second-generation neural networks, termed as Artificial Neural Networks (ANNs), rely on simple static non-linear neurons that are known to be energy-intensive, hindering their implementations on energy-limited processors such as mobile devices. The sparse event-based characteristics of SNNs for information transmission and encoding have made them more feasible for highly energy-efficient neuromorphic computing architectures. The most existing training algorithms for SNNs are based …

Supercapacitors With Gate Electrodes, Tazima Selim Chowdhury

Dissertations

A new approach to improve the capacitance of supercapacitors (SC) is proposed in this study. A typical SC is composed of an anode and a cathode; a separator in between them assures an unintentional discharge of the capacitor. The study focuses on a family of structured separators, either electronically active or passive which are called gates. An active structured separator layer has been fabricated and analyzed. The structured separator has characteristics of electrical diode and is fabricated out of functionalized carbon nanotubes (CNT). Improvement of the overall capacitance of SC, equipped with either active or passive structured separators demonstrated a …

Dual Modality Optical Coherence Tomography : Technology Development And Biomedical Applications, Farzana Rahmat Zaki

Dissertations

Optical coherence tomography (OCT) is a cross-sectional imaging modality that is widely used in clinical ophthalmology and interventional cardiology. It is highly promising for in situ characterization of tumor tissues. OCT has high spatial resolution and high imaging speed to assist clinical decision making in real-time.

OCT can be used in both structural imaging and mechanical characterization. Malignant tumor tissue alters morphology. Additionally, structural OCT imaging has limited tissue differentiation capability because of the complex and noisy nature of the OCT signal. Moreover, the contrast of structural OCT signal derived from tissue’s light scattering properties has little chemical specificity. Hence, …

High-Performance Learning Systems Using Low-Precision Nanoscale Devices, Nandakumar Sasidharan Rajalekshmi

Dissertations

Brain-inspired computation promises a paradigm shift in information processing, both in terms of its parallel processing architecture and the ability to learn to tackle problems deemed unsolvable by traditional algorithmic approaches. The computational capability of the human brain is believed to stem from an interconnected network of 100 billion compute nodes (neurons) that interact with each other through approximately 10¹⁵ adjustable memory junctions (synapses). The conductance of synapses is modifiable allowing the network to learn and perform various cognitive functions. Artificial neural networks inspired by this architecture have demonstrated even super-human performance in many complex tasks.

Computational systems based …

Epitaxial Growth Of Iii-Nitride Nanostructures And Their Optoelectronic Applications, Moab Rajan Philip

Dissertations

Light-emitting diodes (LEDs) using III-nitride nanowire heterostructures have been intensively studied as promising candidates for future phosphor-free solid-state lighting and full-color displays. Compared to conventional GaN-based planar LEDs, III-nitride nanowire LEDs exhibit numerous advantages including greatly reduced dislocation densities, polarization fields, and quantum-confined Stark effect due to the effective lateral stress relaxation, promising high efficiency full-color LEDs. Beside these advantages, however, several factors have been identified as the limiting factors for further enhancing the nanowire LED quantum efficiency and light output power. Some of the most probable causes have been identified as due to the lack of carrier confinement in …

Surface-Enhanced Raman Detection Of Glucose On Several Novel Substrates For Biosensing Applications, Laila Saad Alqarni

Dissertations

The small normal Raman cross-section of glucose is considered to be a major challenge for its detection by Surface Enhanced Raman Spectroscopy (SERS) for medical applications. These applications include blood glucose level monitoring of diabetic patients and evaluation of patients with other medical conditions, since glucose is a marker for many human diseases. This dissertation focuses on Surface-Enhanced Raman Scattering primarily for the detection of glucose. Some experiments also are carried out for the detection of the corresponding enzyme glucose oxidase that is used in electrochemical glucose sensors and in biofuel cells. This project explores the possibility of utilizing Surface …

Workload Allocation In Mobile Edge Computing Empowered Internet Of Things, Qiang Fan

Dissertations

In the past few years, a tremendous number of smart devices and objects, such as smart phones, wearable devices, industrial and utility components, are equipped with sensors to sense the real-time physical information from the environment. Hence, Internet of Things (IoT) is introduced, where various smart devices are connected with each other via the internet and empowered with data analytics. Owing to the high volume and fast velocity of data streams generated by IoT devices, the cloud that can provision flexible and efficient computing resources is employed as a smart "brain" to process and store the big data generated from …

Dual-Battery Empowered Green Cellular Networks, Xilong Liu

Dissertations

With awareness of the potential harmful effects to the environment and climate change, on-grid brown energy consumption of information and communications technology (ICT) has drawn much attention. Cellular base stations (BSs) are among the major energy guzzlers in ICT, and their contributions to the global carbon emissions increase sustainedly. It is essential to leverage green energy to power BSs to reduce their on-grid brown energy consumption. However, in order to furthest save on-grid brown energy and decrease the on-grid brown energy electricity expenses, most existing green energy related works only pursue to maximize the green energy utilization while compromising the …