Naval Information Warfare Center - Atlantic, Charleston, SC

Laboratory Coordinator:
Josette Francis
P.O. Box 190022
North Charleston, SC 29419-9022
josette.j.francis.civ@us.navy.mil

All participants must be U.S. citizens. NIWC ATLANTIC will not accept Permanent Residents or Dual Citizens as applicants for the ONR Summer Faculty Program. All selected faculty will participate for ten continuous weeks on site from June - August

Research Opportunities for SFRP 2023:

1: Title: Adversarial Machine Learning (ML) and Verification and Validation of ML Models
POC: Jamie Lyle, jamie.r.lyle.civ@us.navy.mil

Alt POC: Luke Overbey, lucas.a.overbey.civ@us.navy.mil

Task Description:  The desired research addresses the development of mathematical and automated tools to better understand and apply adversarial machine learning (ML) or other techniques for verification and validation of machine learning models. These approaches could involve dealing with black box (unknown model characteristics), white box (known model characteristics), or gray box (some known characteristics) ML models. In particular, we are interested in ML models aimed at electro-optical (EO), infrared (IR), and radiofrequency detection and classification problems. Efforts can be related to understanding robustness and sensitivity of ML models to adversarial attacks, changing environmental conditions, and enhancing explainability. This project will focus on one such approach approach — as indicated by the area of expertise of the professor approach — and utilize these tools to provide insight into the efficacy of neural network or other ML model approaches.

Place of Performance: Naval Information Warfare Center Atlantic, Charleston, SC
Number of Professors Desired:  1

 

2: Title: Transitioning from Simulations to Real World Scenarios with Reinforcement Learning (RL) Models
POC: Jeff Richley, jeffrey.e.richley2.civ@us.navy.mil

Alt POC: Luke Overbey, lucas.a.overbey.civ@us.navy.mil

Task Description:  The desired research is focused on understanding challenges and mitigation measures to address the transition from low or medium fidelity simulation models to real world experiments when training and evaluating reinforcement learning (RL) models. Specifically, we are interested in enhancing realism in simulation-based training measures that improve performance as models are applied to autonomous scenarios. We will focus research in wargaming (red vs. blue) RL modeling efforts with applications toward maritime autonomous systems. This project will focus on one such approach — as indicated by the area of expertise of the professor  — and utilize this research to provide insight into the application of RL models to realistic/real-world scenarios.

Place of Performance: Naval Information Warfare Center Atlantic, Charleston, SC
Number of Professors Desired:  1

 

3: Title: Machine Learning (ML) and Mathematical Model Development and Research for Decision Tools

POC: Darleen Perez-Lavin, darleen.s.perez-lavin.civ@us.navy.mil

Alt POC: Luke Overbey, lucas.a.overbey.civ@us.navy.mil

Task Description:  The desired research is aimed at understanding and applying game theoretics, reinforcement learning, network theory, path planning, and optimization methods for developing algorithms to create and/or recommend efficient decision tools or plans for logistics, resupply, maintenance scheduling, or other operations research (OR) related applications. We are also interested in understanding tradeoffs between multiple dependent and/or independent objectives. We are interested in both mathematical theory and efficient and/or effective model development research. This project will focus on one such approach — as indicated by the area of expertise of the professor — and utilize tools, models, or algorithms to provide insight into the strategic and tactical decision-making problems with naval applications.

Place of Performance: Naval Information Warfare Center Atlantic, Charleston, SC
Number of Professors Desired:  1

 

4: Title: Utilization of Answer Set Programming as a Decision Tool for Naval Applications

POC: Patrick Kahl, patrick.t.kahl.civ@us.navy.mil

Alt POC: Luke Overbey, lucas.a.overbey.civ@us.navy.mil

Task Description:  The desired research is aimed at understanding and applying answer set programming (ASP) to produce modular agents that can react to event patterns to trigger appropriate associated actions/behaviors. Given such a set of agents, we wish to develop a framework around this that can be utilized as a decision tool that can be utilized toward applications related to naval logistics, resupply, operations research, and command and control applications. This project will focus on one such approach — as indicated by the area of expertise of the professor — and utilize this research and development to provide insight into explainable decision-making tools for naval applications.

Place of Performance: Naval Information Warfare Center Atlantic, Charleston, SC
Number of Professors Desired:  1

 

5: Title: Neuromorphic Computing and Dynamic Vision Sensor (DVS) Cameras

POC: Bryan Ek, bryan.t.ek.civ@us.navy.mil

Alt POC: Luke Overbey, lucas.a.overbey.civ@us.navy.mil

Task Description:  The desired research is aimed at understanding and applying neuromorphic computing algorithms and hardware (e.g. Intel Loihi chips), specifically with an aim of applications that utilize spiking neural networks (SNN) for computer vision (CV) based machine learning (ML) applications. We are also interested in connecting neuromorphic computing and hardware research with the utilization of DVS cameras for efficient detection of movement or object tracking. This project will focus on either hardware or software research or a combination of the two — as indicated by the area of expertise of the professor — and utilize this research to provide insight into efficient edge processing hardware/software solutions for Naval, Marine Corps, and security applications.

Place of Performance: Naval Information Warfare Center Atlantic, Charleston, SC
Number of Professors Desired:  1

 

6: Title: Modulation Classification for Radiofrequency Machine Learning (RF ML)

POC: Georgianna Campbell, georgianna.l.campbell.civ@us.navy.mil

Alt POC: Luke Overbey, lucas.a.overbey.civ@us.navy.mil

Task Description:  The desired research is aimed at understanding and applying novel machine learning (ML) methods to radiofrequency (RF) signals for performing modulation classification. Toward these goals, we are particularly interested in research ideas around low size weight and power (SWAP) hardware/algorithm solutions for edge computing, as well as methods for characterizing quality of algorithms relative to hardware, software, or other constraints. This project will focus on either hardware or software research or a combination of the two — as indicated by the area of expertise of the professor — and utilize this research to provide insight into efficient RF ML and edge processing hardware/software solutions for Naval, Marine Corps, and security applications.

Place of Performance: Naval Information Warfare Center Atlantic, Charleston, SC
Number of Professors Desired:  1

 

7: Title: Cyber Vulnerability Analysis of Mobile and Embedded Systems and Kernels
POC: Gregory Ross, gregory.j.ross.civ@us.navy.mil

Task Description:  The desired research addresses the creation of methods, techniques, individual tools, and sequencing that speed up and automate the discovery of potential vulnerabilities with a low false positive rate or confidence scoring system. The research focus on discovering potential vulnerabilities in mobile and embedded systems with an emphasis is on kernels, as opposed to user space applications, and without access to the source code not for the system being analyzed. We seek knowledge of the state of the art in vulnerability analysis and look to leverage subject matter expertise to develop new tools and tool chains that can be used to independently test and improve cyber defense using new analysis techniques.

Place of performance: Naval Information Warfare Center Atlantic, Charleston, SC
Number of Professors Desired:  1

 

8: Title: Rapid Development of Cyber POCs

POC: Gregory Ross, gregory.j.ross.civ@us.navy.mil

Task Description:  The desired research addresses the development of methods, techniques, and tools to rapidly develop proof of concept code for a given CVE on a given device. The proof of concept is used to test and validate the existence of a vulnerability on a given system. Doing so rapidly is crucial as the number of devices increase in mobile and IOT ecosystems. Tools or techniques to develop new POCs from CVEs or port existing POCs to new devices would assist in the speed of testing new devices for known vulnerabilities.

Place of performance: Naval Information Warfare Center Atlantic, Charleston, SC
Number of Professors Desired:  1

 

9: Title: Online Privacy; Device Data Collection and Device to User Association

POC: Gregory Ross, gregory.j.ross.civ@us.navy.mil

Task Description: The desired research investigates the collection and usage of device and user data across different types of devices. The research focuses on (what) identifying the data being collected, (who) access required to collect, (how) methods of collection, the relative value of identifiers for tracking and association, the techniques used to associate users to and across devices, how these data types can be used to create a user profile. We seek to understand methods and techniques leveraging mobile computing, big data, and machine learning are being used to identify devices, profile a user, associate a user profile to the identity of the user of a device, and track the user across all of their devices.

Place of performance: Naval Information Warfare Center Atlantic, Charleston, SC
Number of Professors Desired:  1

 

10: Title: Identification of Fake Users and Accounts Used By Troll Farms

POC: Gregory Ross, gregory.j.ross.civ@us.navy.mil

Task Description: The desired research addresses the methods and data needed to identify fake users on online platforms. The research focuses on the surveying existing and developing new methods and techniques used to identify personas and user accounts that are fraudulent. It should not only look at user details or activity but also look at identifiers gathered from the devices used to access the platform. We seek knowledge of the state of the art in online user validation and look to leverage subject matter expertise to develop methods using online and device data for the identification of fake accounts.

Place of performance: Naval Information Warfare Center Atlantic, Charleston, SC
Number of Professors Desired:  1