Open Access. Powered by Scholars. Published by Universities.®

Computer Sciences Commons™

Open Access. Powered by Scholars. Published by Universities.®

Air Force Institute of Technology

Discipline
Keyword
Publication Year
Publication
Publication Type
File Type

Articles 121 - 150 of 1277

Full-Text Articles in Computer Sciences

Distributed Reconnaissance Deception Using Software-Defined Networking In A Dynamic Network Environment, Richard Hunter Feustel Mar 2023

Distributed Reconnaissance Deception Using Software-Defined Networking In A Dynamic Network Environment, Richard Hunter Feustel

Theses and Dissertations

This research outlines the design and implementation of a DRDS, which is a RDS distributed across multiple controllers that is capable of deploying reconnaissance deception across multiple switches to mitigate network enumeration by a compromised host. This research outlines the design and development of the DRDS as well as tests its functional abilities and routing performance when compared to a two other network routing solutions: a legacy network solution and centralized ONOS controller scheme deploying layer 2 forwarding. The functional tests proved the system can properly route traffic across 100% of the tested scenarios carrying traffic that includes IP, ARP, …

Go to article

Intel Total Memory Encryption: Functional Verification And Performance Analysis, Tallas T. S. Goo Mar 2023

Intel Total Memory Encryption: Functional Verification And Performance Analysis, Tallas T. S. Goo

Theses and Dissertations

While more attention is generally focused on software security, computer hardware security remains an important effort. Should an attacker gain direct physical access, computers with little to no hardware security can quickly be compromised via a manner of methods. One such attacker method is to steal information directly from the active memory of a locked, powered-on computer. To counter this attack, a hardware security method was developed called memory encryption. Memory encryption, as the name suggests, protects against adversary methods like cold boot attacks by encrypting all of memory. This research evaluates the efficacy and performance specifically of Intel TME. …

Go to article

The Electromagnetic Bayonet: Development Of A Scientific Computing Method For Aperture Antenna Optimization, Michael P. Ingold Mar 2023

The Electromagnetic Bayonet: Development Of A Scientific Computing Method For Aperture Antenna Optimization, Michael P. Ingold

Theses and Dissertations

The quiet zone of a radar range is the region over which a transmitted EM field approximates a uniform plane wave to within some finite error tolerance. Any target to be measured must physically fit within this quiet zone to prevent excess measurement error. Compact radar ranges offer significant operational advantages for performing RCS measurements but their quiet zone sizes are constrained by space limitations. In this work, a scientific computing approach is used to investigate whether equivalent-current transmitters can be designed that generate larger quiet zones than a conventional version at short range. A time-domain near-field solver, JefimenkoModels, was …

Go to article

Fast And Accurate 3d Object Reconstruction For Cargo Load Planning, Adam R. Nasi Mar 2023

Fast And Accurate 3d Object Reconstruction For Cargo Load Planning, Adam R. Nasi

Theses and Dissertations

Cargo load planning involves efficiently packing objects into aircraft subject to constraints such as space and weight distribution. Currently, this is performed manually by loadmasters. The United States Air Force is investigating ways to automate this process in order to improve airlift operational readiness while saving money. The first step in such a process would be generating 3D reconstructions of cargo objects to be used by a load planning algorithm. To that end, this thesis presents a novel method for fast, scaled, and accurate 3D reconstruction of cargo objects. This method can scan a 2.5m×3m×2m object in less than 10 …

Go to article

Characterizing Location-Based Electromagnetic Leakage Of Computing Devices Using Convolutional Neural Networks To Increase The Effectiveness Of Side-Channel Analysis Attacks, Ian C. Heffron Mar 2023

Characterizing Location-Based Electromagnetic Leakage Of Computing Devices Using Convolutional Neural Networks To Increase The Effectiveness Of Side-Channel Analysis Attacks, Ian C. Heffron

Theses and Dissertations

SCA attacks aim to recover some sort of secret information, often in the form of a cipher key, from a target device. Some of these attacks focus on either power-based leakage, or EM-based leakage. Neural networks have recently gained in popularity as tools in SCA attacks. Near-field EM probes with high-spatial resolution enable attackers to isolate physical locations above a processor. This enables attackers to exploit the spatial dependencies of algorithms running on said processor. These spatial dependencies result in different physical locations above a chip emanating different signal strengths. The strengths of different locations can be mapped using the …

Go to article

Adaptation Of Network Flow Problems For Course Of Action Generation, Alexander N. Stephens Mar 2023

Adaptation Of Network Flow Problems For Course Of Action Generation, Alexander N. Stephens

Theses and Dissertations

This thesis introduces two methods to generate Courses of Action (COA) in distributed warfare scenarios: the Wargaming Commodity Course of Action Automated Method Under Uncertainty (WCCAAM-U2) and Dynamic Transshipment Problem (DTP)-generated COAs. Previous work by Deberry et al. used a Multi-Commodity Flow Problem (MCFP) to generate COAs for single-period wargame scenarios with known enemy force amounts. In WCCAAM-U2, we adapt an MCFP to work in situations where only intelligence estimates of enemy forces are known. Compared to two other COA-generation methods, the WCCAAAM-U2 COA outperforms the next highest-performing COA by 307% when compared by a ratio of objective success rate …

Go to article

Spectral Material Classification Of Orbital Objects - Applying Machine Learning To Visible And Near-Infrared Spectral Scenes, Stephen M. Stumpf Mar 2023

Spectral Material Classification Of Orbital Objects - Applying Machine Learning To Visible And Near-Infrared Spectral Scenes, Stephen M. Stumpf

Theses and Dissertations

MSI and HSI techniques allow users to determine the material composition of an object at range. To avoid labor-intensive manual classification, ML is used to determine the most likely material contained in a given pixel of a target image. Previous work primarily focuses on terrestrial applications; this paper extends these techniques into the low-illumination space situational awareness domain, which is of critical importance to national security. HSI datacubes are preprocessed with RL deconvolution as a means of reducing the effects of the optical PSF; then, statistical ML techniques, including k-NN, LDA, QDA, and SVMs are implemented as means of assigning …

Go to article

Ensemble Aggregation In A Multi-Perspective Environment, Jonathan P. Nash Mar 2023

Ensemble Aggregation In A Multi-Perspective Environment, Jonathan P. Nash

Theses and Dissertations

Research towards improving the performance of artificial intelligence networks has found that larger and more complex networks tends to yield better results, and continuous hardware upgrades enables the development of larger, more complicated, and better performing neural networks. However, many devices that are widely available and more practical to everyday use, such as drones or smartphones, are unable to use the state-of-the-art neural networks because they simply do not have the processing capabilities to run them in addition to their normal function. It is possible to overcome this lower performance by using a variety of these smaller neural networks as …

Go to article

Classifying Open-Air Target Measurements Using Simulation-Trained Convolutional Neural Networks, Matthew M. Rofrano Mar 2023

Classifying Open-Air Target Measurements Using Simulation-Trained Convolutional Neural Networks, Matthew M. Rofrano

Theses and Dissertations

This research focuses on the development of machine learning networks that can identify and classify airborne targets using their radar cross section response. Simulation and measurement data for five targets was collected using Altair's CadFEKO software, and the Air Force Institute's Compact Radar Range. Three machine learning models were trained using simulation data, and evaluated using the collected measurement data. Variability is introduced to the training data by applying random gaussian noise to simulation results. Gaussian noise is added to the measurement data prior to evaluation in-order to model "hostile noise jamming." Network performance is measured against a baseline performance …

Go to article

Monocular Vision And Machine Learning For Pose Estimation, Quang Ngoc Tran Mar 2023

Monocular Vision And Machine Learning For Pose Estimation, Quang Ngoc Tran

Theses and Dissertations

This thesis introduces a monocular vision-based approach for 6 DoF pose estimation on a known object. The proposed solution is to use a CNN to find known features of an object in an image. These known features, together with their known locations, are used by a PnP algorithm to estimate the pose of the target object with respect to the camera. The primary difficulty with CNN-based methods is needing to generate a large amount of training data to effectively create the CNN. To overcome this difficulty, a 3D model of the real-world object is created and used in a visualization …

Go to article

Detection Algorithms And Clutter Metrics Comparison For Long Wave Infrared Point-Source Targets, Rudolf N. Vonniederhausern Mar 2023

Detection Algorithms And Clutter Metrics Comparison For Long Wave Infrared Point-Source Targets, Rudolf N. Vonniederhausern

Theses and Dissertations

In Infrared Search and Track (IRST) systems, clutter in the image hinders target detection especially in point-source target scenarios. Currently there is not a standardized metric for quantifying background clutter. Many clutter metrics have been proposed, but none have demonstrated effectiveness or compatibility for point-source targets. Factors such as environment conditions, detection algorithm, and correlation coefficient to probability of detection (PD) and false alarm (PFA) are the main considerations in determining the effectiveness of clutter metrics. Determining the most successful metric will increase Air Force Test and Evaluation (T&E) units’ capability by providing additional information on test conditions and environments …

Go to article

Safe And Reliable Software And The Formal Verification Of Prim's Algorithm In Spark, Brian S. Wheelhouse Mar 2023

Safe And Reliable Software And The Formal Verification Of Prim's Algorithm In Spark, Brian S. Wheelhouse

Theses and Dissertations

Despite evidence that formal verification helps produce highly reliable and secure code, formal methods, i.e., mathematically based tools and approaches for software and hardware verification, are not commonly used in software and hardware development. The limited emphasis on formal verification in software education and training suggests that many developers have never considered the benefits of formal verification. Despite the challenging nature of their mathematical roots, software verification tools have improved; making it easier than ever to verify software. SPARK, a programming language and a formal verification toolset, is of particular interest for the AFRL, and will be a primary focus …

Go to article

Effectiveness Of A Timing Side Channel For Deriving Neural Network Depth, Matthew P. Weeks Mar 2023

Effectiveness Of A Timing Side Channel For Deriving Neural Network Depth, Matthew P. Weeks

Theses and Dissertations

From facial recognition on cell phones to vehicle traffic modeling for city planning, integrating ML models can be an expensive investment in resources. Protecting that investment is difficult, as information about the model and how it was built can be leaked through multiple channels, such as timing and memory access. In this thesis, one method of extracting data through a timing side-channel is examined across multiple hardware and software configurations to determine its reliability for general use. While attempting to determine the layer count of a target model solely from its inference time, the research determined that it is not …

Go to article

Strategic Action Execution Through Regret Matching In Press Diplomacy, Leif D. White Mar 2023

Strategic Action Execution Through Regret Matching In Press Diplomacy, Leif D. White

Theses and Dissertations

To take most advantage of collaboration, negotiation is paramount to succeed in press Diplomacy. Humans use this construct to work towards self victory or sometimes towards an alternative strategic objective undefined in the game’s rules. To emulate this behavior, this thesis examines how to use communication to enable the victory or defeat of any other player in the game. This research develops a press Diplomacy agent, Lyre, that can work to attain these specific objectives in Diplomacy through the regret matching algorithm (RM). We also study how Lyre can begin Diplomacy with the goal to win, then shift strategies to …

Go to article

Using Embedded Systems And Augmented Reality For Automated Aerial Refueling, Nathaniel A. Wilson Mar 2023

Using Embedded Systems And Augmented Reality For Automated Aerial Refueling, Nathaniel A. Wilson

Theses and Dissertations

The goal of automated aerial refueling (AAR) is to extend the range of unmanned aircraft. Control latency prevents a human from remotely controlling the receiving aircraft as it approaches a tanker. To conform with the size, weight, and power constraints of a small unmanned aircraft, an AAR system must execute in real-time on an embedded platform. This thesis explores the timing and computational performance of a NVIDIA Jetson AGX Orin to a state-of-the-art general-purpose computer using existing AAR algorithms. It also constructs an augmented reality framework as an intermediate step for testing vision-based AAR algorithms between virtual testing and expensive …

Go to article

Temporal Convolutional Neural Networks For Device Discrimination, Ryan T. Zacher Mar 2023

Temporal Convolutional Neural Networks For Device Discrimination, Ryan T. Zacher

Theses and Dissertations

This research uses TCN modifications to CNN classifiers, specifically dilation, causal padding, and residual blocks, to focus on temporal features and improve existing DNA analysis processes. Dilation significantly improves classification accuracy, even detecting features where no other models were able to. The smallest improvement shown is a 3-6dB reduction in SNR to reach a 90% classification accuracy. The maximum improvement is shown in the data-delivery region of the Cisco dataset, with the dilated model being the only model to exceed 90% classification accuracy. The other TCN modifications are shown to have no beneficial effect on the models.

Go to article

An Analysis Of Cloud Computing Migration Costs And Effects For Dod Applications, Joseph S. Moore Iv Mar 2023

An Analysis Of Cloud Computing Migration Costs And Effects For Dod Applications, Joseph S. Moore Iv

Theses and Dissertations

The Air Force launched “Cloud One” in 2017. Cloud One provides cloud computing options for military applications. Cloud One provides common secure computing environments, standardized platforms, application migration and support services, and data management. Currently, Cloud One has over one hundred mission applications on board. Although there is information on the cost, performance, personnel requirements, risks, and migration of the commercial sector and cloud options, there is limited recorded information on the same topics for Cloud One. As such, there is a gap in the literature regarding data/feedback for mission applications that have migrated to Cloud One. This research takes …

Go to article

Garbage In ≠ Garbage Out: Exploring Gan Resilience To Image Training Set Degradations, Nicholas M. Crino Mar 2023

Garbage In ≠ Garbage Out: Exploring Gan Resilience To Image Training Set Degradations, Nicholas M. Crino

Theses and Dissertations

Generative Adversarial Networks (GANs) have received increasing attention in recent years due to their ability to capture complex, high-dimensional data distributions without the need for extensive labeling. Since their conception in 2014, a wide array of GAN variants have been proposed featuring alternative architectures, optimizers, and loss functions with the goal of improving performance and training stability. While this research has yielded GAN variants robust to training set shrinkage and corruption, our research focuses on quantifying the resilience of a GAN architecture to specific modes of image degradation. We conduct systematic experimentation to determine empirically the effects of 10 fundamental …

Go to article

Classification And Analysis Of Twitter Bot And Troll Accounts, Callan P. Mccormick Mar 2023

Classification And Analysis Of Twitter Bot And Troll Accounts, Callan P. Mccormick

Theses and Dissertations

This research trains, tests, and analyzes bot and troll classification models using publicly available, open source datasets. Specifically, it applies decision tree, random forest, feed forward neural networks, and long-short term memory neural networks with hyperparameters tuned via designed experiment to five labeled bot datasets created between 2011 and 2020 and one dataset labeling state-sponsored disinformation accounts or trolls. The first three models utilize account profile features, while the last model applies natural language processing techniques, specifically GloVe embedding, to analyze a user’s Tweet history. Results indicate that the random forest model outperforms the other three models with an average …

Go to article

Bayesian Recurrent Neural Networks For Real Time Object Detection, Stephen Z. Kimatian Mar 2023

Bayesian Recurrent Neural Networks For Real Time Object Detection, Stephen Z. Kimatian

Theses and Dissertations

Neural networks have become increasingly popular in real time object detection algorithms. A major concern with these algorithms is their ability to quantify their own uncertainty, leading to many high profile failures. This research proposes three novel real time detection algorithms. The first of leveraging Bayesian convolutional neural layers producing a predictive distribution, the second leveraging predictions from previous frames, and the third model combining these two techniques together. These augmentations seek to mitigate the calibration problem of modern detection algorithms. These three models are compared to the state of the art YOLO architecture; with the strongest contending model achieving …

Go to article

Hierarchical Federated Learning On Healthcare Data: An Application To Parkinson's Disease, Brandon J. Harvill Mar 2023

Hierarchical Federated Learning On Healthcare Data: An Application To Parkinson's Disease, Brandon J. Harvill

Theses and Dissertations

Federated learning (FL) is a budding machine learning (ML) technique that seeks to keep sensitive data private, while overcoming the difficulties of Big Data. Specifically, FL trains machine learning models over a distributed network of devices, while keeping the data local to each device. We apply FL to a Parkinson’s Disease (PD) telemonitoring dataset where physiological data is gathered from various modalities to determine the PD severity level in patients. We seek to optimally combine the information across multiple modalities to assess the accuracy of our FL approach, and compare to traditional ”centralized” statistical and deep learning models.

Go to article

Air Force Cadet To Career Field Matching Problem, Ian P. Macdonald Mar 2023

Air Force Cadet To Career Field Matching Problem, Ian P. Macdonald

Theses and Dissertations

This research examines the Cadet to Air Force Specialty Code (AFSC) Matching Problem (CAMP). Currently, the matching problem occurs annually at the Air Force Personnel Center (AFPC) using an integer program and value focused thinking approach. This paper presents a novel method to match cadets with AFSCs using a generalized structure of the Hospitals Residents problem with special emphasis on lower quotas. This paper also examines the United States Army Matching problem and compares it to the techniques and constraints applied to solve the CAMP. The research culminates in the presentation of three algorithms created to solve the CAMP and …

Go to article

Predicting Success Of Pilot Training Candidates Using Interpretable Machine Learning, Alexandra S. King Mar 2023

Predicting Success Of Pilot Training Candidates Using Interpretable Machine Learning, Alexandra S. King

Theses and Dissertations

The United States Air Force (USAF) has struggled with a sustained pilot shortage over the past several years; senior military and government leaders have been working towards a solution to the problem, with no noticeable improvements. Both attrition of more experienced pilots as well as wash out rates within pilot training contribute to this issue. This research focuses on pilot training attrition. Improving the process for selecting pilot candidates can reduce the number of candidates who fail. This research uses historical specialized undergraduate pilot training (SUPT) data and leverages select machine learning techniques to determine which factors are associated with …

Go to article

Simulation And Analysis Of Dynamic Threat Avoidance Routing In An Anti-Access Area Denial (A2ad) Environment, Dante C. Reid Mar 2023

Simulation And Analysis Of Dynamic Threat Avoidance Routing In An Anti-Access Area Denial (A2ad) Environment, Dante C. Reid

Theses and Dissertations

This research modeled and analyzed the effectiveness of different routing algorithms for penetration assets in an A2AD environment. AFSIM was used with different configurations of SAMs locations and numbers to compare the performance of AFSIM’s internal zone and shrink algorithm routers with a Dijkstra algorithm router. Route performance was analyzed through computational and operational metrics, including computational complexity, run-time, mission survivability, and simulation duration. This research also analyzed the impact of the penetration asset’s ingress altitude on those factors. Additionally, an excursion was conducted to analyze the Dijkstra algorithm router’s grid density holding altitude constant to understand its impact on …

Go to article

Jtag-Based Extraction Of Processor And Memory Data For Anomaly Detection In Cyber-Physical Systems, Jonathan D. Price Mar 2023

Jtag-Based Extraction Of Processor And Memory Data For Anomaly Detection In Cyber-Physical Systems, Jonathan D. Price

Theses and Dissertations

This document is a compilation of two scholarly articles that together compose a graduate thesis. The first article assesses the viability of commercially available JTAG debuggers to access runtime memory and register data, finding that commercial JTAG debuggers are not suitable for runtime data extraction for anomaly detection in cyber-physical systems due to slow memory and register access times. Because of this gap in technology, the second article designs, implements, and tests a novel, custom-built architecture named JTAG Data Extraction Tool (J-DET) for program counter-sampling. J-DET is the first device purpose-built to extract processor information in runtime for attack detection; …

Go to article

A Reinforcement Learning Approach To A Beyond Visual Range Air Combat Maneuvering Problem, Caleb A. Taylor Mar 2023

A Reinforcement Learning Approach To A Beyond Visual Range Air Combat Maneuvering Problem, Caleb A. Taylor

Theses and Dissertations

A one-versus-one air combat maneuvering problem is considered wherein a friendly autonomous aircraft must engage and defeat an adversary autonomous aircraft in a beyond visual range environment. The Advanced Framework for Simulation, Integration, and Modeling (AFSIM) is leveraged to model the complex and interdependent operations of aircraft, sensors, and weapons utilized in beyond visual range air combat. We formulate a Markov decision process to obtain high-quality decision policies wherein our autonomous aircraft makes maneuvering and missile firing decisions. We utilize a reinforcement learning solution procedure that implements a linear value function approximation to represent state-decision pairs due to the high …

Go to article

Air Force Digital Badges, Jacob Chan Mar 2023

Air Force Digital Badges, Jacob Chan

Theses and Dissertations

The Air Force talent management and force development systems are antiquated. Airmen records are often stored on different Air Force information systems. Existing records sometimes lack granularity and context to recognize Airmen skills. Digital badges are a newer technology utilized by academia and industry to recognize member skills. However, military badging research is sparse and existing studies do not provide sufficient evidence on the value of digital badging to the Air Force. The studies: (1) lack background research on badging; (2) do not provide quantitative data on the effects of badging; and (3) issued badges through commercial entities which may …

Go to article

Cellphone-Acoustics Based Suas Detection And Tracking, Ryan D. Clendening Mar 2023

Cellphone-Acoustics Based Suas Detection And Tracking, Ryan D. Clendening

Theses and Dissertations

Small Unmanned Aerial Systems (sUAS) are an easily accessible technology that has become an increasingly large threat to US critical systems. This threatening technology demands using fault-tolerant, low-cost, replaceable, and accurate sensing resources, which counter the ubiquitous nature of sUAS [1]. Therefore, the methods developed in this thesis detect and track sUAS using easily accessible sensing resources, such as cellphones. First, we develop an acoustics sensor network-based sUAS detection methodology. In the latter effort, a deep learning model is trained using the acoustics data from the data collection to predict sUAS range from a cellphone. Combined, these two efforts demonstrate …

Go to article

Entering Hyperspace: Conditional Hyperspectral Reflectance Image Generation Using Convolutional Neural Networks, Bret M. Wagner Mar 2023

Entering Hyperspace: Conditional Hyperspectral Reflectance Image Generation Using Convolutional Neural Networks, Bret M. Wagner

Theses and Dissertations

The field of remote sensing continues to expand in both commercial and defense domains. Development of advanced space based EOIR sensors has driven corresponding demand for sensor data for algorithm development. The AFIT Sensor and Scene Emulation Tool (ASSET) produces realistic synthetic electro-optical and infrared (EO/IR) data with absolute truth for the purpose of clutter suppression, target detection, and tracking algorithm development. This thesis presents a novel model which transforms panchromatic images into realistic hyperspectral reflectance images. The direct application of this model is to allows users to generate hyperspectral background images as inputs to ASSET allowing users to benefit …

Go to article

Automated Registration Of Titanium Metal Imaging Of Aircraft Components Using Deep Learning Techniques, Nathan A. Johnston Mar 2023

Automated Registration Of Titanium Metal Imaging Of Aircraft Components Using Deep Learning Techniques, Nathan A. Johnston

Theses and Dissertations

Studies have shown a connection between early catastrophic engine failures with microtexture regions (MTRs) of a specific size and orientation on the titanium metal engine components. The MTRs can be identified through the use of Electron Backscatter Diffraction (EBSD) however doing so is costly and requires destruction of the metal component being tested. A new methodology of characterizing MTRs is needed to properly evaluate the reliability of engine components on live aircraft. The Air Force Research Lab Materials Directorate (AFRL/RX) proposed a solution of supplementing EBSD with two non-destructive modalities, Eddy Current Testing (ECT) and Scanning Acoustic Microscopy (SAM). Doing …

Go to article

First Prev Next Last