Computer Engineering Commons^™

Promise And Limitations Of Supervised Optimal Transport-Based Graph Summarization Via Information Theoretic Measures, Sepideh Neshatfar

Electronic Theses and Dissertations

Graph summarization is a fundamental problem in the field of data analysis, aiming to distill extensive graph datasets into more manageable, yet informative representations. The challenge lies in creating compressed graphs that faithfully retain crucial structural information for downstream tasks. A recent advancement in this domain introduces an optimal transport-based framework that enables the incorporation of a priori knowledge regarding the importance of nodes, edges, and attributes during the graph summarization process. However, the statistical properties of this innovative framework remain largely unexplored. This master's thesis embarks on a comprehensive exploration of the field of graph summarization, with a particular …

Modeling The Impact Of The Covid-19 Pandemic On Speeding At Rural Roadway Facilities In Maine Using Short-Term Speed And Traffic Count Data., Amirhossein Shahlaeegilan

Electronic Theses and Dissertations

The COVID-19 pandemic caused a significant change in traffic operations and safety. For instance, various U.S. states reported an increase in the rate of fatal and severe injury crashes over this duration. In April and May 2020, the comprehensive stay-at-home orders were issued across the country, including in Maine. These orders resulted in drastic reductions in traffic volume,switching working policies of noticeable number of corporations and educational administrations like universities to the remote working, closure of many organizations during the lockdown period, and people avoidance to public places to suppress the possible exposure to the virus were among the most …

Comparison Between Cpu And Gpu For Parallel Implementation For A Neural Network Model Using Tensorflow And A Big Dataset, Intisar Alkaabwi

Electronic Theses and Dissertations

Machine learning is a rapidly growing field that has become more common of late. Because of the demanding computational usage of machine learning, this field has many dimensions needing research. TensorFlow has been developed to deal with and analyze neural networks computation. In particular, TensorFlow is often used in one of the machine learning branches and is called deep learning. This work discusses the performance of a deep learning model to train a very large dataset with TensorFlow. It compares performance when the run happens on CPUs and on GPUs regarding the run time and speed. The run time is …

Evaluating Tlb (Translation Lookaside Buffer) Performance Overhead For Nvm (Non-Volatile Memory) Hybrid System, Xiang Guo

Electronic Theses and Dissertations

As the non-volatile memory (NVM) technology offers near-DRAM performance and near-disk capacity, NVM has emerged as a new storage class. Conventional file systems, designed for hard disk drives or solid-state drives, need to be re-examined or even re-designed for NVM storage. For example, new file systems such as NOVA, HMFS, HMVFS and Ext4-DAX, have been developed and implemented to fully leverage NVM’s characteristics, such as fast fine-grained access. This thesis research uses a variety of I/O workloads to evaluate the performance overhead of the TLB (translation lookaside buffer) in various file systems on emulated NVM storage systems, in which NVM …

How Gpu Rendering Affects Image Processing And Scientific Calculation Speed, Power And Energy On A Raspberry Pi, Qihao He

Electronic Theses and Dissertations

In this thesis, we explore the speed, power, and energy performance of the same data process on the central processing unit (CPU) with and without the acceleration of the Graphics Processing Unit (GPU) on the microcomputer Raspberry Pi (RPI). We tested on the RPI in two different fields. The first was comparing the speed, power, and energy usage with and without GPU acceleration in the image processing impacts on RPI model B+. The second was comparing speed, power, energy usage, and accuracy for scientific calculation with and without GPU acceleration on RPI model B+ and 3B.

We used a novel …

Detailed Power Measurement With Arm Embedded Boards, Yanxiang Mao

Electronic Theses and Dissertations

Power and energy are becoming important considerations in today's electronic equipment. The amount of power required to run a supercomputer for an hour could supply an ordinary household for many months. The need for low-power computing also extends to smaller devices, such as mobile phones, laptops and embedded devices.

In order to optimize power usage of electronic equipment, we need to collect information on the power consumption of these devices. Unfortunately it is not easy to do this on modern computing systems. Existing measuring equipment is often expensive, inaccurate, and difficult to operate. The main goal of this project is …

Integrated Environment And Proximity Sensing For Uav Applications, Shawn S. Brackett

Electronic Theses and Dissertations

As Unmanned Aerial Vehicle (UAV), or “drone” applications expand, new methods for sensing, navigating and avoiding obstacles need to be developed. The project applies an Extended Kalman Filter (EKF) to a simulated quadcopter vehicle though Matlab in order to estimate not only the vehicle state but the world state around the vehicle. The EKF integrates multiple sensor readings from range sensors, IMU sensors, and radiation sensors and combines this information to optimize state estimates. The result is an estimated world map to be used in vehicle navigation and obstacle avoidance.

The simulation handles the physics behind the vehicle flight. As …

Spatial Relations And Natural-Language Semantics For Indoor Scenes, Stacy A. Doore

Electronic Theses and Dissertations

Over the past 15 years, there have been increased efforts to represent and communicate spatial information about entities within indoor environments. Automated annotation of information about indoor environments is needed for natural-language processing tasks, such as spatially anchoring events, tracking objects in motion, scene descriptions, and interpretation of thematic places in relationship to confirmed locations. Descriptions of indoor scenes often require a fine granularity of spatial information about the meaning of natural-language spatial utterances to improve human-computer interactions and applications for the retrieval of spatial information. The development needs of these systems provide a rationale as to why—despite an extensive …

Accelerating Scientific Computing Models Using Gpu Processing, Raymond F. Flagg Iii

Electronic Theses and Dissertations

GPGPUs offer significant computational power for programmers to leverage. This computational power is especially useful when utilized for accelerating scientific models. This thesis analyzes the utilization of GPGPU programming to accelerate scientific computing models.

First the construction of hardware for visualization and computation of scientific models is discussed. Several factors in the construction of the machines focus on the performance impacts related to scientific modeling.

Image processing is an embarrassingly parallel problem well suited for GPGPU acceleration. An image processing library was developed to show the processes of recognizing embarrassingly parallel problems and serves as an excellent example of converting …

Detailed Low-Cost Energy And Power Monitoring Of Computing Systems, Chad M. Paradis

Electronic Theses and Dissertations

Power and energy are increasingly important metrics in modern computing systems. Large supercomputers utilize millions of cores and can consume as much power as a small town; monitoring and reducing power consumption is an important task. At the other extreme, power usage of embedded and mobile devices is also critically important. Battery life is a key concern in such devices; having detailed power measurement allows optimizing these devices for power as well. Current systems are not set up to allow easy power measurement. There has been much work in this area, but obtaining power readings is often expensive, intrusive, and …

Challenges Of Implementing Automatic Dependent Surveillance Broadcast In The Nextgen Air Traffic Management System, Carl J. Giannatto Jr.

Electronic Theses and Dissertations

The Federal Aviation Administration is in the process of replacing the current Air Traffic Management (ATM) system with a new system known as NextGen. Automatic Dependent Surveillance-Broadcast (ADS-B) is the aircraft surveillance protocol currently being introduced as a part of the NextGen system deployment. The evolution of ADS-B spans more than two decades, with development focused primarily on increasing the capacity of the Air Traffic Control (ATC) system and reducing operational costs. Security of the ADS-B communications network has not been a high priority, and the inherent lack of security measures in the ADS-B protocol has come under increasing scrutiny …

Towards A Filmic Look And Feel In Real Time Computer Graphics, Sherief Farouk

Electronic Theses and Dissertations

Film footage has a distinct look and feel that audience can instantly recognize, making its replication desirable for computer generated graphics. This thesis presents methods capable of replicating significant portions of the film look and feel while being able to fit within the constraints imposed by real-time computer generated graphics on consumer hardware.

Improving The Performance Of The Parallel Ice Sheet Model On A Large-Scale, Distributed Supercomputer, Timothy J. Morey

Electronic Theses and Dissertations

In this thesis, we describe our work to understand and improve the performance and scalability of the Parallel Ice Sheet Model (PISM) on the Ranger supercomputer. PISM enables the simulation of large-scale ice sheets, such as those found in Greenland and Antarctica, which are of particular interest to climate scientists due to their potential to contribute to sea-level rise.

PISM has a unique parallel architecture that is designed to take advantage of the computational resources available on state-of-the-art supercomputers. The problem, however, is that even though PISM can run without modifcation on a supercomputer, it is generally unable to do …

Assessment Of Audio Interfaces For Use In Smartphone Based Spatial Learning Systems For The Blind, Shreyans Jain

Electronic Theses and Dissertations

Recent advancements in the field of indoor positioning and mobile computing promise development of smart phone based indoor navigation systems. Currently, the preliminary implementations of such systems only use visual interfaces—meaning that they are inaccessible to blind and low vision users. According to the World Health Organization, about 39 million people in the world are blind. This necessitates the need for development and evaluation of non-visual interfaces for indoor navigation systems that support safe and efficient spatial learning and navigation behavior. This thesis research has empirically evaluated several different approaches through which spatial information about the environment can be conveyed …

Mobile Map Browsers: Anticipated User Interaction For Data Pre-Fetching, Benjamin T. Weber

Electronic Theses and Dissertations

When browsing a graphical display of geospatial data on mobile devices, users typically change the displayed maps by panning, zooming in and out, or rotating the device. Limited storage space on mobile devices and slow wireless communications, however, impede the performance of these operations. To overcome the bottleneck that all map data to be displayed on the mobile device need to be downloaded on demand, this thesis investigates how anticipated user interactions affect intelligent pre-fetching so that an on-demand download session is extended incrementally. User interaction is defined as a set of map operations that each have corresponding effects on …

High-Speed Digital And Mixed-Signal Components For X– And Ku–Band Direct Digital Synthesizers In Indium Phosphide Dhbt Technology, Steven Eugene Turner

Electronic Theses and Dissertations

Recently reported double heterojunction bipolar transistor (DHBT) devices manufactured in Indium Phosphide (InP) technology with ft and fmax both over 300 GHz enable advanced high-speed digital and mixed-signal circuits. In this thesis, the use of InP DHBT devices for high-speed accumulator circuits and X– and Ku–band direct digital synthesizer (DDS) circuits are investigated. At these frequencies, new technological challenges in the design of digital and mixed-signal circuits arise in areas including power consumption and clock distribution. This thesis addresses the speed/power tradeoffs in high-speed accumulator designs, the design of DDS circuits, and clock distribution simulation. The results of six accumulator …

Remote Sensing And Data Collection In A Marine Science Application, Isaac Abraham Horn

Electronic Theses and Dissertations

In every field of science, research is driven by the data scjentists are able to collect. Specifically in the area of marine science, researchers at the Darling Marine Center (DMC) in Walpole, Maine study thc emergence events of undersea organisms. This thesis describes the development of a system integrating hardware infrastructure and existing software technology with a new software application, for the purpose of data acquisition at the DMC. In their research, a Tracor Acoustic Profiling System (TAPS), measures the water column density by utilizing sonar pulscs at several different frequencies. A serial interface provides access to the systems command …

Gene Expression Data Analysis Using Fuzzy Logic, Robert Reynolds

Electronic Theses and Dissertations

DNA microarray technology allows for the parallel analysis of the expression of genes in an organism. The wealth of spatio-temporal data provided by the technology allows us to attempt to reverse engineer the genetic network. Fuzzy logic has been proposed as a method of analyzing the relationships between genes as well as their corresponding proteins. Combinations of genes are entered into a fuzzy model of gene interaction and evaluated on the basis of how well the combination fits the model. Those combinations of genes that fit the model are likely to be related. However, current analysis algorithms are slow and …