Naval Postgraduate School
Calhoun, Institutional Archive of the Naval Postgraduate SchoolNot a member yet
71232 research outputs found
Sort by
OBFUSCATION, STEALTH, AND NON-ATTRIBUTION IN RED TEAM TOOLS
No full textCurrent automated red teaming tools are limited in their ability to emulate advanced persistent threat (APT) behaviors. Supporting such behaviors in automated security assessments and tools can be helpful for improving organizations’ cyber defense preparedness. This research enhances the Cyber Automated Red Team Tool (CARTT) by integrating advanced evasion techniques to better simulate sophisticated cyber threats. By incorporating Metasploit Framework evasion modules and new custom Internet Control Message Protocol (ICMP) and Domain Name System (DNS) evasion capabilities into CARTT, its ability to evade detection by common security controls is significantly improved. The research demonstrates how obfuscation, stealth, and non-attribution techniques can be effectively automated into red teaming tools. The enhanced CARTT was tested in a simulated operational environment, which demonstrated its effectiveness in identifying vulnerabilities and assessing the robustness of security measures. Results of the research showed successful evasion of antivirus detection systems and covert data exfiltration using the newly implemented techniques. The enhanced CARTT will enable network managers as well as cybersecurity professionals to conduct more thorough evaluations of defense mechanisms against sophisticated threats, ultimately strengthening overall cybersecurity postures.Distribution Statement A. Approved for public release: Distribution is unlimited.Outstanding ThesisLieutenant, United States Nav
USMC Manpower Models Modernization II
No full textThis project builds on our previous work in support of an FY22 Naval Research Program project (Seagren et al., 2022). The primary objective of this project is to determine the feasibility and effectiveness of a more automated process to more accurately account for all active-duty personnel gains and losses by month and facilitate the inclusion of this information in various models and processes. We develop a machine learning version of the Enlisted End-Strength Planning Model (ESPM) and compare its performance to the legacy process. Ultimately, we find that it is possible to develop and implement a machine learning approach that effectively forecasts base NEAS attrition in a managerially useful manner. In fact, evidence suggests that the machine learning models may dramatically outperform the legacy methods and that these findings are robust across fiscal years. However, resources required to develop the machine learning models include significant time on a High-Performance Computer, which may not be practical for the topic sponsor. While we find that the machine learning models perform well, the benefits relative to the legacy model and process likely do not justify the effort given current constraints of computing power.Approved for public release; distribution is unlimited.This research is supported by funding from the Naval Postgraduate School, Naval Research Program (PE0605853N/2098). https://nps.edu/nrpChief of Naval Operations (CNO)Naval Postgraduate School, Naval Research Program, Headquarters Marine Corps, Manpower & Reserve Affair
STATISTICAL ANALYSIS OF MANPOWER ATTRITION IN THE NAVAL RESERVE OFFICER TRAINING CORPS
Get PDFThe Naval Reserve Officer Training Corps (NROTC) program is essential for developing future leaders for the United States Navy and Marine Corps. This study aims to analyze and understand the factors contributing to midshipman attrition within the NROTC program, utilizing a dataset of over 5,000 midshipmen who enrolled and subsequently left the program between 2012 and the present. Statistical techniques and machine learning algorithms, including logistic regression, random forests, gradient boosting machines, K-Nearest Neighbors (KNN), and Naive Bayes, were used to identify key predictors of attrition. The analysis showed higher dropout rates at certain institutions and during specific periods, underscoring the need for targeted interventions and time-informed counseling strategies. Gender differences were also noted, with females more likely to leave due to physical reasons and males due to academic and disciplinary issues, suggesting the importance of gender-specific support systems. Although machine learning models showed moderate improvements in predictive accuracy, their effectiveness was limited, in part due to imbalanced data and the complexity of human factors.Distribution Statement A. Approved for public release: Distribution is unlimited.Lieutenant Junior Grade, United States Nav
Water Submersion Detection Switch
Get PDFA water submersion detection system that may include a buoyant device having a conductive surface, and a housing enclosing the buoyant device and having conductive elements. The conductive elements may include a first set of one or more nonadjacent conductive elements that are electrically connected, and a second set of one or more nonadjacent conductive elements that are electrically connected. The system may include a submersion alert device that activates responsive to the buoyant device contacting the housing
LOCAL DESCRIPTION OF FRICTIONAL WEAKENING IN VIBRATED GRANULAR SHEAR FLOWS
Get PDFThe rheology of vibrated granular shear flows is a key missing piece for modeling and predicting landslides, earthquakes, and more. In this thesis, we use computer simulations written in the Large-scale Atomic/Molecular Multiprocessor Parallel Simulator (LAMMPS) language to study frictional properties of vibrated granular shear flows subjected to externally applied vibrations at the lower boundary. Based on previous dimensional analysis and the Discrete Element Method (DEM) simulations from LAMMPS, we examine frictional weakening at the local grain-scale, correlating stress, contact fabric, and force fabric tensors with fluctuations in grain velocities and system friction . This thesis makes progress toward an end goal of writing down a closed constitutive law for vibrated granular shear flows that could solve for more complex geometries that currently can only be solved with nonlocal models. We gratefully acknowledge funding by the Army Research Office (grants W911NF1510012 and W911NF2220044).Distribution Statement A. Approved for public release: Distribution is unlimited.Outstanding ThesisLieutenant, United States Nav
IMPLEMENTATION OF DECENTRALIZED UNMANNED SURFACE VEHICLE SWARM CONTROL
No full textAlgorithms for swarm control of unmanned surface vessels (USV) normally require a communication network utilizing radio frequency (RF) to determine range and bearing between vehicles to maintain formation. The RF signals, either between USVs or between each USV and a centralized swarm controller, are vulnerable to detection and disruption by enemy actions, thus preventing typical methods of swarm formation control. This research decentralizes an existing control algorithm by simulating USVs with computer vision capable of determining range and bearing to fellow USVs in the swarm. Performance of the decentralized control algorithm will be analyzed to determine the impact of simulated cameras with different fields of view, as well as different desired swarm behaviors.Distribution Statement A. Approved for public release: Distribution is unlimited.Lieutenant, United States Nav
PANDEMIC PREPAREDNESS FOR PHOENIX REGIONAL EMS
Get PDFThe COVID-19 global pandemic has brought heightened awareness to the challenges that biological events can pose for emergency medical service providers. This research seeks to produce policy guidance to create optimal pandemic preparedness strategies for the Phoenix Fire Department and its regional partners. Utilizing a design thinking process, the research engaged focus groups in a series of sessions designed to understand and define pandemic response challenges, leading to possible solutions. Focus group findings revealed that the department faced multifaceted issues in pandemic preparation, ranging from equipment and personnel decontamination to inter-agency support coordination. However, the Phoenix Fire Department also demonstrated robust capabilities in confronting past challenges. The inherent unpredictability and severity of pandemics necessitate preemptive preparation, anchored by the prioritization of accountability and thorough planning. It is recommended that the Phoenix Fire Department establish a specialized pandemic preparedness group dedicated to drafting and updating a comprehensive pandemic preparedness plan. Incorporating a continual review mechanism, this plan would benefit from a pandemic preparedness audit tool, ensuring iterative improvements through regular testing, integration of the latest knowledge, and lessons learned.Distribution Statement A. Approved for public release: Distribution is unlimited.Civilian, Phoenix Fire Departmen
AN EMPIRICAL ANALYSIS TO DETERMINE A COMMON AUGMENTED REALITY OR MIXED REALITY SOLUTION TO IMPROVE TRAINING AND OPERATIONAL CAPABILITIES FOR THE MARINE CORPS
Get PDFI Marine Expeditionary Force sponsored research to identify methods of developing applications for various augmented reality (AR), virtual reality (VR), and mixed reality (XR) use cases. The research sought to identify methods of hand and arm tracking that would assist with the completion of tasks. This thesis provides the background for the growing demand in the Marine Corps for the use of AR/VR/XR systems grounded in Project Tripoli. The method of thesis research identifies various AR/VR/XR systems that assists in better understanding current capabilities in commercially developed systems. Site visits were conducted to various laboratories to learn about the systems and to learn about a new technology under development by the Defense Advanced Research Projects Agency (DARPA). Perceptually enabled task guidance (PTG) is an artificial intelligence solution to providing task guidance to the user to assist with the completion of various tasks. PTG uses computer vision (CV) technology to identify actions taken by the user and provides detailed instructions for the user to complete the task. Instructions are provided to the user through an AR/XR head-mounted device. AR/XR displays overlay data and projects the data in front of the user where it can be referenced in real time without the need to look away at a laptop, publication, or reference material. AR/XR supported by AI can support multiple military use cases across a wide range of military communities.Distribution Statement A. Approved for public release: Distribution is unlimited.Outstanding ThesisMajor, United States Marine CorpsAcquisition Research Program, Monterrey, California 93943NPS Naval Research ProgramThis project was funded in part by the NPS Naval Research Program
MASTERING THE DIGITAL ART OF WAR: DEVELOPING INTELLIGENT COMBAT SIMULATION AGENTS FOR WARGAMING USING HIERARCHICAL REINFORCEMENT LEARNING
No full textIn today’s rapidly evolving military landscape, advancing artificial intelligence (AI) in support of wargaming becomes essential. Despite reinforcement learning (RL) showing promise for developing intelligent agents, conventional RL faces limitations in handling the complexity inherent in combat simulations. This dissertation proposes a comprehensive approach, including targeted observation abstractions, multi-model integration, a hybrid AI framework, and an overarching hierarchical reinforcement learning (HRL) framework. Our localized observation abstraction using piecewise linear spatial decay simplifies the RL problem, enhancing computational efficiency and demonstrating superior efficacy over traditional global observation methods. Our multi-model framework combines various AI methodologies, optimizing performance while still enabling the use of diverse, specialized individual behavior models. Our hybrid AI framework synergizes RL with scripted agents, leveraging RL for high-level decisions and scripted agents for lower-level tasks, enhancing adaptability, reliability, and performance. Our HRL architecture and training framework decomposes complex problems into manageable subproblems, aligning with military decision-making structures. Although initial tests did not show improved performance, insights were gained to improve future iterations. This study underscores AI’s potential to revolutionize wargaming, emphasizing the need for continued research in this domain.Distribution Statement A. Approved for public release: Distribution is unlimited.Lieutenant Colonel, United States Marine Corp