Engineering Commons^™

Efficient End-To-End Autonomous Driving, Hesham Eraqi

Theses and Dissertations

Steering a car through traffic is a complex task that is difficult to cast into algorithms. Therefore, researchers turn to train artificial neural networks from front-facing camera data stream along with the associated steering angles. Nevertheless, most existing solutions consider only the visual camera frames as input, thus ignoring the temporal relationship between frames. In this work, we propose a Convolution Long Short-Term Memory Recurrent Neural Network (C-LSTM), which is end-to-end trainable, to learn both visual and dynamic temporal dependencies of driving. Additionally, We introduce posing the steering angle regression problem as classification while imposing a spatial relationship between the …

Framework For Reasoning With Speech Processing, Ahmed Laarfi

Theses and Dissertations

It is known that programming languages are textual. We try here to take the advantages of Speech Recognition (SR) and employ them in creating a verbal Language, which takes its instruction from the voice. Because this work is a novel approach in the programming world, we could not find any resources. This dissertation aims to make essential developments in Speech Recognition (SR)and Artificial Intelligence by constructing a new compiler that receives commands verbally and executes them. That means entering data into the Computer by voice commands. This method of input means that we link several major computer topics with several …

Modern Standard Arabic Speech Recognition: Using Formants Measurements To Extract Vowels From Arabic Words’ Consonant-Vowel-Consonant-Vowel Structure, Mohamed Ali Alshaari

Theses and Dissertations

Arabic texts suffer from missing diacritics (short vowels) which become obstacles for new learners. Speech Recognition is the translation of words spoken to text through intelligent computer programs. As of today, it has been integrated into many computer systems. Arabic Speech Recognition has made progress over the years, but it is still not as good as English speech recognition due to the problem of short vowels not being recognized. This is mainly because the Arabic language is unlike the English language in the nature because it is a Semitic language. This is reflected in different characteristics such as grammar, morphology, …

Cloud Computing Service Interoperability And Architectural Concepts, Anmar Salih

Theses and Dissertations

Today’s market with a vast number of Cloud-computing providers creates a challenge for practical cooperation between the various provider cloud service platforms. Not only does Cloud interoperability provide this needed cooperation, but it also avoids vendor lock-in and, additionally, saves time and cost. Although there is no established definition for Cloud interoperability, most researchers agree on the purposes of Cloud interoperability. Research and literature have attempted to explain interoperability as transferring data, moving workloads, and migrating virtual machines between Cloud platforms. Transferring data between Clouds refers to objects migrating between provider-specific domains. In comparison, data migration is the most commonly …

Efficient Edge Analytics: Addressing Cyber-Physical Masint With Machine Learning On Audio At The Edge, David Elliott

Theses and Dissertations

With the growth of the Internet of Things and the rise of Big Data, data processing and machine learning applications are being moved to cheap and low size, weight, and power (SWaP) devices at the edge, often in the form of mobile phones, embedded systems, or microcontrollers. The field of Cyber-Physical Measurements and Signature Intelligence (MASINT) makes use of these devices to analyze and exploit data in ways not otherwise possible, which results in increased data quality, increased security, and decreased bandwidth. However, methods to train and deploy models at the edge are limited, and models with sufficient accuracy are …

Correlation Between Air Quality Index And Traffic Volume Using Internet Of Things (Iot), Omar Sayah Alruwaili

Theses and Dissertations

In highly populated world areas, such as metropolises, hazardous air pollution has been linked to the presence of damaging climate and health issues, which are becoming increasingly common. Indeed, a major problem facing urban areas today is air pollution. Gas emissions from vehicles can be seen as the most important source of this kind of pollution. Pollutant gases emitted as parts of car exhaust consist of chemicals such as carbon monoxide (CO), nitrogen dioxide (NO2), and sulphur dioxide (SO2), and ozone (O3), as well as particulate matter (PM). In some places in the world, non-governmental and foreign corporations have also …

Dbknot: A Transparent And Seamless, Pluggable Tamper Evident Database, Islam Khalil

Theses and Dissertations

Database integrity is crucial to organizations that rely on databases of important data. They suffer from the vulnerability to internal fraud. Database tampering by internal malicious employees with high technical authorization to their infrastructure or even compromised by externals is one of the important attack vectors.

This thesis addresses such challenge in a class of problems where data is appended only and is immutable. Examples of operations where data does not change is a) financial institutions (banks, accounting systems, stock market, etc., b) registries and notary systems where important data is kept but is never subject to change, and c) …

Direct Digital Synthesis: A Flexible Architecture For Advanced Signals Research For Future Satellite Navigation Payloads, Pranav R. Patel

Theses and Dissertations

In legacy Global Positioning System (GPS) Satellite Navigation (SatNav) payloads, the architecture does not provide the flexibility to adapt to changing circumstances and environments. GPS SatNav payloads have largely remained unchanged since the system became fully operational in April 1995. Since then, the use of GPS has become ubiquitous in our day-to-day lives. GPS availability is now a basic assumption for distributed infrastructure; it has become inextricably tied to our national power grids, cellular networks, and global financial systems. Emerging advancements of easy to use radio technologies, such as software-defined radios (SDRs), have greatly lowered the difficulty of discovery and …

Joint 1d And 2d Neural Networks For Automatic Modulation Recognition, Luis M. Rosario Morel

Theses and Dissertations

The digital communication and radar community has recently manifested more interest in using data-driven approaches for tasks such as modulation recognition, channel estimation and distortion correction. In this research we seek to apply an object detector for parameter estimation to perform waveform separation in the time and frequency domain prior to classification. This enables the full automation of detecting and classifying simultaneously occurring waveforms. We leverage a lD ResNet implemented by O'Shea et al. in [1] and the YOLO v3 object detector designed by Redmon et al. in [2]. We conducted an in depth study of the performance of these …

Performance Analysis Of V2v And V2i Communications Using Empirical Path Loss Models Indicators And Embedded Iot Devices, Ibrahim Lateef Oraibi Al Kinoon

Theses and Dissertations

Vehicle management technologies deals with the management of critical vehicle information, including location, idle time, speed, and mileage. Such information can always be transferred through a direct vehicle-to-vehicle communication among cars. However, the limitation of this type of design is that it is based on the assumption that vehicles are always served by cellular bases, which is not always the case. For the effective implementation of the Internet-of-Things (IoT) technology in this sector, it is critical to design vehicles with systems that enable them to transmit essential information in the absence of base stations. IoT technologies can then be used …

One Dimensional Neural Time Series Generation, Kaleb Earl Smith

Theses and Dissertations

Time dependent data is a main source of information in today’s data driven world.Generating this type of data though has shown its challenges and made it an interesting research area in the field of generative machine learning. The challenge with this one-dimensional (1D) data has been for applications in machine learning to gain access to a considerable amount of quality data needed for algorithm development and analysis. Modeling synthetic data using a Generative Adversarial Network (GAN) has been at the heart of providing a viable solution. Our work focuses on one dimensional times series and explores the “few shot” generation …

Two Techniques For Automated Logging Statement Evolution, Allan R. Spektor

Theses and Dissertations

This thesis presents and explores two techniques for automated logging statement evolution. The first technique reinvigorates logging statement levels to reduce information overload using degree of interest obtained via software repository mining. The second technique converts legacy method calls to deferred execution to achieve performance gains, eliminating unnecessary evaluation overhead.

On The Robustness Of Bayesian Network Learning Algorithms Against Malicious Attacks, Noah Joseph Geveke

Theses and Dissertations

Bayesian networks are effective tools for discovering relationships between variables in a data set. Algorithms that learn Bayesian networks from data fall into three categories: constraint-based, score-based, and hybrid. Hybrid algorithms contain a constraint testing sub-procedure as well as a score function to create the network. Malicious changes to the training set can cause invalid networks that do not model the true data. The effects of these changes have been demonstrated using the PC algorithm, a constraint-based algorithm. In this thesis a method was developed to measure the robustness of various algorithms to determine potential malicious changes. The robustness analysis …

Evaluating And Improving The Seu Reliability Of Artificial Neural Networks Implemented In Sram-Based Fpgas With Tmr, Brittany Michelle Wilson

Theses and Dissertations

Artificial neural networks (ANNs) are used in many types of computing applications. Traditionally, ANNs have been implemented in software, executing on CPUs and even GPUs, which capitalize on the parallelizable nature of ANNs. More recently, FPGAs have become a target platform for ANN implementations due to their relatively low cost, low power, and flexibility. Some safety-critical applications could benefit from ANNs, but these applications require a certain level of reliability. SRAM-based FPGAs are sensitive to single-event upsets (SEUs), which can lead to faults and errors in execution. However there are techniques that can mask such SEUs and thereby improve the …

Dynamic Reconfigurable Real-Time Video Processing Pipelines On Sram-Based Fpgas, Andrew Elbert Wilson

Theses and Dissertations

For applications such as live video processing, there is a high demand for high performance and low latency solutions. The configurable logic in FPGAs allows for custom hardware to be tailored to a specific video application. These FPGA designs require technical expertise and lengthy implementation times by vendor tools for each unique solution. This thesis presents a dynamically configurable topology as an FPGA overlay to deploy custom hardware processing pipelines during run-time by utilizing dynamic partial reconfiguration. Within the FPGA overlay, a configurable topology with a routable switch allows video streams to be copied and mixed to create complex data …

Revisiting Elliptic Curve Cryptography With Applications To Post-Quantum Sidh Ciphers, Wesam Nabil Eid

Theses and Dissertations

Elliptic Curve Cryptography (ECC) has positioned itself as one of the most promising candidates for various applications since its introduction by Miller and Kolbitz in 1985 [53, 44]. The core operation for ECC is the scalar multiplication [k]P where many efforts have addressed its computation speed. Here we introduce an efficient approach for calculating elliptic curve operations by a novel regrouping of terms and creating new projective representation operators and increasing parallelism. These operators and the corresponding projective coordinate representations are shown to lead to adjusted versions of scalar multiplication algorithms that are evaluated. These techniques enable more opportunities for …

On The Characterization Of Natural Language Structure And Literary Stylometry - A Network Science Approach, Younis Anas Younis Al Rozz

Theses and Dissertations

Natural language processing (NLP) techniques have been through many advancements in recent years, linguistics and scientist utilized these techniques to solve many challenges related to written language and literary. Problems such as finding the genetic relationships among languages, attributing author of a text and categorizing text by genre have been treated throughout the years using conventional statistical methods, for instance, bag of words (BoW), N-gram, the frequency of words and the lexical distance between words. By considering written language as a complex system, network science tools and techniques can be used to address those problems. A unified methodology is proposed …

Geolocation Based On Signal Level Measurement And Time Advance Inside The Network, Zaenab Shakir

Theses and Dissertations

The increasing demands for location-based services in recent years led to provide accurate location information inside the network. The applications for location-based service are assisting emergency request, fraud defense, social media, and marketing. All these demands make the position accuracy highly required. Various techniques and methods have been developed to estimate the position of mobile device inside the network such as RSS, TOA, AOA, and TDOA. In this work, a new method to estimate the accuracy of locating active LTE cellular subscribers. The proposed method is a network-based technique and relies on the Reference Signal Received Power (RSRP) measurements and …

Model Optimization For Edge Devices, Adolf Anthony D’Costa

Theses and Dissertations

Edge devices are undergoing groundbreaking computing transformation, which lets us tap into artificial intelligence, quantum computing, 5th generation network capability, fog networking, and computing complex algorithms. Edge systems have substantial advantages over the conventional system in terms of scalability, optimized resources, reliability, and security. The proliferation of such resource-constrained devices in recent years has resulted in the generation of a large quantity of data; these data-producing devices are attractive targets for applications of machine learning. Machine learning models, especially deep learning neural networks, produced models that have high accuracy and prediction capability, but it comes at the cost of computation …

Improving Wake-Up-Word And General Speech Recognition Systems, Gamal Mohamed Bohouta

Theses and Dissertations

Automatic Speech Recognition (ASR), a technology that allows a machine to recognize the utterances spoken into a microphone by a person and then converts it to text, is commonly used for different types of applications, such as command and control systems, personal assistant systems, medical systems, disabilities systems, dictation systems, telephony systems, and embedded applications. Due to its extensive use, interest in ASR technology has surged among inventors and researchers alike. They have worked diligently to improve the performance of the ASR systems by developing several techniques or approaches in different aspects,such as enhancing features, training an acoustic model, enhancing …

An Overlay Architecture For Pattern Matching, Rasha Elham Karakchi

Theses and Dissertations

Deterministic and Non-deterministic Finite Automata (DFA and NFA) comprise the fundamental unit of work for many emerging big data applications, motivating recent efforts to develop Domain-Specific Architectures (DSAs) to exploit fine-grain parallelism available in automata workloads.

This dissertation presents NAPOLY (Non-Deterministic Automata Processor Over- LaY), an overlay architecture and associated software that attempt to maximally exploit on-chip memory parallelism for NFA evaluation. In order to avoid an upper bound in NFA size that commonly affects prior efforts, NAPOLY is optimized for runtime reconfiguration, allowing for full reconfiguration in 10s of microseconds. NAPOLY is also parameterizable, allowing for offline generation of …

Parsimonious Sociology Theory Construction: From A Computational Framework To Semantic-Based Parsimony Analysis, Mingzhe Du

Theses and Dissertations

In the social sciences, theories are used to explain and predict observed phenomena in the natural world. Theory construction is the research process of building testable scientific theories to explain and predict observed phenomena in the natural world. Conceptual new ideas and meanings of theories are conveyed through carefully chosen definitions and terms.

The principle of parsimony, an important criterion for evaluating the quality of theories (e.g., as exemplified by Occam’s Razor), mandates that we minimize the number of definitions (terms) used in a given theory.

Conventional methods for theory construction and parsimony analysis are based on heuristic approaches. However, …

A Machine Learning Based Approach To Accelerate Catalyst Discovery, Asif Jamil Chowdhury

Theses and Dissertations

Computational catalysis, in contrast to experimental catalysis, uses approximations such as density functional theory (DFT) to compute properties of reaction intermediates. But DFT calculations for a large number of surface species on variety of active site models are resource intensive. In this work, we are building a machine learning based predictive framework for adsorption energies of intermediate species, which can reduce the computational overhead significantly. Our work includes the study and development of appropriate machine learning models and effective fingerprints or descriptors to predict energies accurately for different scenarios. Furthermore, Bayesian inverse problem, that integrates experimental catalysis with its computational …

Multi-Channel Security Through Data Fragmentation, Micah J. Hayden

Theses and Dissertations

This thesis presents a novel security system developed for a multi-channel communication architecture, which achieves security by distributing the message and its associated message authentication code across the available channels at the bit level, to support systems that require protection from confidentiality and integrity attacks without relying solely on traditional encryption. One contribution of the work is to establish some helpful terminology, present a basic theory for multi-channel communications, describe the services provided by an optimal system, and then implement a proof of concept system to demonstrate the concept's validity. This proof of concept, focused on the splitting and recombination …

A Comparative Evaluation Of The Detection And Tracking Capability Between Novel Event-Based And Conventional Frame-Based Sensors, James P. Boettiger

Theses and Dissertations

Traditional frame-based technology continues to suffer from motion blur, low dynamic range, speed limitations and high data storage requirements. Event-based sensors offer a potential solution to these challenges. This research centers around a comparative assessment of frame and event-based object detection and tracking. A basic frame-based algorithm is used to compare against two different event-based algorithms. First event-based pseudo-frames were parsed through standard frame-based algorithms and secondly, target tracks were constructed directly from filtered events. The findings show there is significant value in pursuing the technology further.

Developing A Serious Game To Explore Joint All Domain Command And Control, Nathaniel W. Flack

Theses and Dissertations

Changes in the geopolitical landscape and increasing technological complexity have prompted the U.S. Military to coin Multi-Domain Operations (MDO) and Joint All-Domain Command and Control as terms to describe an over-arching strategy that frames the complexity of warfare across both traditional and emerging warfighting domains. Teaching new and advanced concepts associated with these terms requires both innovation as well as distinct education and training tools in order to realize the cultural change advocated by senior military leaders. BSN, a Collectible Card Game, was developed to teach concepts integral to MDO and initiate discussion on military strategy.

Sliver: Simulation-Based Logic Bomb Identification/Verification For Unmanned Aerial Vehicles, Jake M. Magness

Theses and Dissertations

This research introduces SLIVer, a Simulation-based Logic Bomb Identification/Verification methodology, for finding logic bombs hidden within Unmanned Aerial Vehicle (UAV) autopilot code without having access to the device source code. Effectiveness is demonstrated by executing a series of test missions within a high-fidelity software-in-the-loop (SITL) simulator. In the event that a logic bomb is not detected, this methodology defines safe operating areas for UAVs to ensure to a high degree of confidence the UAV operates normally on the defined flight plan. SLIVer uses preplanned flight paths as the baseline input space, greatly reducing the input space that must be searched …

Implications And Limitations Of Securing An Infiniband Network, Lucas E. Mireles

Theses and Dissertations

The InfiniBand Architecture is one of the leading network interconnects used in high performance computing, delivering very high bandwidth and low latency. As the popularity of InfiniBand increases, the possibility for new InfiniBand applications arise outside the domain of high performance computing, thereby creating the opportunity for new security risks. In this work, new security questions are considered and addressed. The study demonstrates that many common traffic analyzing tools cannot monitor or capture InfiniBand traffic transmitted between two hosts. Due to the kernel bypass nature of InfiniBand, many host-based network security systems cannot be executed on InfiniBand applications. Those that …

A Study Of Execution Performance For Rust-Based Object Vs Data Oriented Architectures, Joseph A. Vagedes

Theses and Dissertations

To investigate the Data-Oriented Design (DOD) paradigm, in particular, an architecture built off its principles: the Entity-Component-System (ECS). ECS is commonly used by video game engines due to its ability to store data in a way that is optimal for the cache to access. Additionally, the structure of this paradigm produces a code-base that is simple to parallelize as the workload can be distributed across a thread-pool based on the data used with little to no need for data safety measures such as mutexes and locks. A final benefit, although not easily measured, is that the DOD paradigm produces a …

Interoperable Ads-B Confidentiality, Brandon C. Burfeind

Theses and Dissertations

The worldwide air traffic infrastructure is in the late stages of transition from legacy transponder systems to Automatic Dependent Surveillance - Broadcast (ADS-B) based systems. ADS-B relies on position information from GNSS and requires aircraft to transmit their identification, state, and position. ADS-B promises the availability of high-fidelity air traffic information; however, position and identification data are not secured via authentication or encryption. This lack of security for ADS-B allows non-participants to observe and collect data on both government and private flight activity. This is a proposal for a lightweight, interoperable ADS-B confidentiality protocol which uses existing format preserving encryption …