Publications

Abstract not provided.

Securing satellite groundstations against cyber-attacks is vital to national security missions. However, these cyber threats are constantly evolving. As vulnerabilities are discovered and patched, new vulnerabilities are discovered and exploited. In order to automate the process of discovering existing vulnerabilities and the means to exploit them, a reinforcement learning framework is presented in this report. We demonstrate that this framework can learn to successfully navigate an unknown network and detect nodes of interest despite the presence of a moving target defense. The agent then exfiltrates a file of interest from the node as quickly as possible. This framework also incorporates a defensive software agent that learns to impede the attacking agents progress. This setup allows for the agents to work against each other and improve their abilities. We anticipate that this capability will help uncover unforeseen vulnerabilities and the means to mitigate them. The modular nature of the framework enables users to swap out learning algorithms and modify the reward functions in order to adapt the learning tasks to various use cases and environments. Several algorithms, viz., tabular Q learning, deep Q networks, proximal policy optimization, advantage actor-critic, generative adversarial imitation learning, are explored for the agents and the results highlighted. The agent learns to solve the tasks in a light-weight abstract environment. Once the agent learns to perform sufficiently well, it can be deployed in a minimega virtual machine environment (or a real network) with wrappers that map abstract actions to software commands. The agent also uses a local representation of the actions called a ‘slot-mechanism’. This allows the agent to learn in a certain network and generalize it to different networks. The defensive agent learns to predict the actions taken by an offensive agent and uses that information to anticipate the threat. This information can then either be used to raise an alarm or to take actions to thwart the attack. We believe that with the appropriate reward design, a representative environment, and action set, this framework can be generalized to tackle other cybersecurity tasks. By sufficiently training these agents, we can anticipate vulnerabilities leading to robust future designs. We can also deploy automated defensive agents that can help secure satellite groundstation and their vital national security missions.

Cyberattacks against industrial control systems have increased over the last decade, making it more critical than ever for system owners to have the tools necessary to understand the cyber resilience of their systems. However, existing tools are often qualitative, subject matter expertise-driven, or highly generic, making thorough, data-driven cyber resilience analysis challenging. The ADROC project proposed to develop a platform to enable efficient, repeatable, data-driven cyber resilience analysis for cyber-physical systems. The approach consists of two phases of modeling: computationally efficient math modeling and high-fidelity emulations. The first phase allows for scenarios of low concern to be quickly filtered out, conserving resources available for analysis. The second phase supports more detailed scenario analysis, which is more predictive of real-world systems. Data extracted from experiments is used to calculate cyber resilience metrics. ADROC then ranks scenarios based on these metrics, enabling prioritization of system resources to improve cyber resilience.

Abstract not provided.

Recent high profile cyber attacks on critical infrastructures have raised awareness about the severe and widespread impacts that these attacks can have on everyday life. This awareness has spurred research into making industrial control systems and other cyber-physical systems more resilient. A plethora of cyber resilience metrics and frameworks have been proposed for cyber resilience assessments, but these approaches typically assume that data required to populate the metrics is readily available, an assumption that is frequently not valid. This paper describes a new cyber experimentation platform that can be used to generate relevant data and to calculate resilience metrics that quantify how resilient specified industrial control systems are to specified threats. Demonstration of the platform and analysis process are illustrated through a use case involving the control system for a pressurized water reactor.

Abstract not provided.

Abstract not provided.

Abstract not provided.

This report summarizes the activities performed as part of the Science and Engineering of Cybersecurity by Uncertainty quantification and Rigorous Experimentation (SECURE) Grand Challenge LDRD project. We provide an overview of the research done in this project, including work on cyber emulation, uncertainty quantification, and optimization. We present examples of integrated analyses performed on two case studies: a network scanning/detection study and a malware command and control study. We highlight the importance of experimental workflows and list references of papers and presentations developed under this project. We outline lessons learned and suggestions for future work.

Abstract not provided.

Abstract not provided.

Concerns about cyber threats to space systems are increasing. Researchers are developing intrusion detection and protection systems to mitigate these threats, but sparsity of cyber threat data poses a significant challenge to these efforts. Development of credible threat data sets are needed to overcome this challenge. This paper describes the extension/development of three data generation algorithms (generative adversarial networks, variational auto-encoders, and generative algorithm for multi-variate timeseries) to generate cyber threat data for space systems. The algorithms are applied to a use case that leverages the NASA Operational Simulation for Small Satellites (NOS$^{3})$ platform. Qualitative and quantitative measures are applied to evaluate the generated data. Strengths and weaknesses of each algorithm are presented, and suggested improvements are provided. For this use case, generative algorithm for multi-variate timeseries performed best according to both qualitative and quantitative measures.

Concerns about cyber threats to space systems are increasing. Researchers are developing intrusion detection and protection systems to mitigate these threats, but sparsity of cyber threat data poses a significant challenge to these efforts. Development of credible threat data sets are needed to overcome this challenge. This paper describes the extension/development of three data generation algorithms (generative adversarial networks, variational auto-encoders, and generative algorithm for multi-variate timeseries) to generate cyber threat data for space systems. The algorithms are applied to a use case that leverages the NASA Operational Simulation for Small Satellites (NOS$^{3})$ platform. Qualitative and quantitative measures are applied to evaluate the generated data. Strengths and weaknesses of each algorithm are presented, and suggested improvements are provided. For this use case, generative algorithm for multi-variate timeseries performed best according to both qualitative and quantitative measures.

Abstract not provided.

Abstract not provided.

Abstract not provided.

Abstract not provided.

Abstract not provided.

Abstract not provided.

Many optical systems are used for specific tasks such as classification. Of these systems, the majority are designed to maximize image quality for human observers. However, machine learning classification algorithms do not require the same data representation used by humans. We investigate the compressive optical systems optimized for a specific machine sensing task. Two compressive optical architectures are examined: an array of prisms and neutral density filters where each prism and neutral density filter pair realizes one datum from an optimized compressive sensing matrix, and another architecture using conventional optics to image the aperture onto the detector, a prism array to divide the aperture, and a pixelated attenuation mask in the intermediate image plane. We discuss the design, simulation, and trade-offs of these systems built for compressed classification of the Modified National Institute of Standards and Technology dataset. Both architectures achieve classification accuracies within 3% of the optimized sensing matrix for compression ranging from 98.85% to 99.87%. The performance of the systems with 98.85% compression were between an F / 2 and F / 4 imaging system in the presence of noise.

Abstract not provided.

Abstract not provided.

Abstract not provided.

Abstract not provided.

Images are often not the optimal data form to perform machine learning tasks such as scene classification. Compressive classification can reduce the size, weight, and power of a system by selecting the minimum information while maximizing classification accuracy.In this work we present designs and simulations of prism arrays which realize sensing matrices using a monolithic element. The sensing matrix is optimized using a neural network architecture to maximize classification accuracy of the MNIST dataset while considering the blurring caused by the size of each prism. Simulated optical hardware performance for a range of prism sizes are reported.

Trust in a microelectronics-based systems can be characterized as the level of confidence that the system is free of subversive alterations inserted by a malicious adversary during system development. Outkin et al. recently developed GPLADD, a game-theoretic framework that enables trust analysis through a set of mathematical models that represent multi-step attack graphs and contention between system attackers and defenders. This paper extends GPLADD to include detection of attacks on development processes and defender decision processes that occur in response to detection events. The paper provides mathematical details for implementing attack detection and demonstrates the models on an example system. The authors further demonstrate how optimal defender strategies vary when solution concepts and objective functions are modified.

Abstract not provided.

Abstract not provided.

Trust in a microelectronics-based system can be characterized as the level of confidence that a system is free of subversive alterations made during system development, or that the development process of a system has not been manipulated by a malicious adversary. Trust in systems has become an increasing concern over the past decade. This article presents a novel game-theoretic framework, called GPLADD (Graph-based Probabilistic Learning Attacker and Dynamic Defender), for analyzing and quantifying system trustworthiness at the end of the development process, through the analysis of risk of development-time system manipulation. GPLADD represents attacks and attacker-defender contests over time. It treats time as an explicit constraint and allows incorporating the informational asymmetries between the attacker and defender into analysis. GPLADD includes an explicit representation of attack steps via multi-step attack graphs, attacker and defender strategies, and player actions at different times. GPLADD allows quantifying the attack success probability over time and the attacker and defender costs based on their capabilities and strategies. This ability to quantify different attacks provides an input for evaluation of trust in the development process. We demonstrate GPLADD on an example attack and its variants. We develop a method for representing success probability for arbitrary attacks and derive an explicit analytic characterization of success probability for a specific attack. We present a numeric Monte Carlo study of a small set of attacks, quantify attack success probabilities, attacker and defender costs, and illustrate the options the defender has for limiting the attack success and improving trust in the development process.

Abstract not provided.

Abstract not provided.

Abstract not provided.

We investigate the feasibility of additively manufacturing optical components to accomplish task-specific classification in a computational imaging device. We report on the design, fabrication, and characterization of a non-traditional optical element that physically realizes an extremely compressed, optimized sensing matrix. The compression is achieved by designing an optical element that only samples the regions of object space most relevant to the classification algorithms, as determined by machine learning algorithms. The design process for the proposed optical element converts the optimal sensing matrix to a refractive surface composed of a minimized set of non-repeating, unique prisms. The optical elements are 3D printed using a Nanoscribe, which uses two-photon polymerization for high-precision printing. We describe the design of several computational imaging prototype elements. We characterize these components, including surface topography, surface roughness, and angle of prism facets of the as-fabricated elements.

Many optical systems are used for specific tasks such as classification. Of these systems, the majority are designed to maximize image quality for human observers; however, machine learning classification algorithms do not require the same data representation used by humans. In this work we investigate compressive optical systems optimized for a specific machine sensing task. Two compressive optical architectures are examined: An array of prisms and neutral density filters where each prism and neutral density filter pair realizes one datum from an optimized compressive sensing matrix, and another architecture using conventional optics to image the aperture onto the detector, a prism array to divide the aperture, and a pixelated attenuation mask in the intermediate image plane. We discuss the design, simulation, and tradeoffs of these compressive imaging systems built for compressed classification of the MNSIT data set. To evaluate the tradeoffs of the two architectures, we present radiometric and raytrace models for each system. Additionally, we investigate the impact of system aberrations on classification accuracy of the system. We compare the performance of these systems over a range of compression. Classification performance, radiometric throughput, and optical design manufacturability are discussed.

Many optical systems are used for specific tasks such as classification. Of these systems, the majority are designed to maximize image quality for human observers; however, machine learning classification algorithms do not require the same data representation used by humans. In this work we investigate compressive optical systems optimized for a specific machine sensing task. Two compressive optical architectures are examined: An array of prisms and neutral density filters where each prism and neutral density filter pair realizes one datum from an optimized compressive sensing matrix, and another architecture using conventional optics to image the aperture onto the detector, a prism array to divide the aperture, and a pixelated attenuation mask in the intermediate image plane. We discuss the design, simulation, and tradeoffs of these compressive imaging systems built for compressed classification of the MNSIT data set. To evaluate the tradeoffs of the two architectures, we present radiometric and raytrace models for each system. Additionally, we investigate the impact of system aberrations on classification accuracy of the system. We compare the performance of these systems over a range of compression. Classification performance, radiometric throughput, and optical design manufacturability are discussed.

Nuisance and false alarms are prevalent in modern physical security systems and often overwhelm the alarm station operators. Deep learning has shown progress in detection and classification tasks, however, it has rarely been implemented as a solution to reduce the nuisance and false alarm rates in a physical security systems. Previous work has shown that transfer learning using a convolutional neural network can provide benefit to physical security systems by achieving high accuracy of physical security targets [10]. We leverage this work by coupling the convolutional neural network, which operates on a frame-by-frame basis, with temporal algorithms which evaluate a sequence of such frames (e.g. video analytics). We discuss several alternatives for performing this temporal analysis, in particular Long Short-Term Memory and Liquid State Machine, and demonstrate their respective value on exemplar physical security videos. We also outline an architecture for developing an ensemble learner which leverages the strength of each individual algorithm in its aggregation. The incorporation of these algorithms into physical security systems creates a new paradigm in which we aim to decrease the volume of nuisance and false alarms in order to allow the alarm station operators to focus on the most relevant threats.

Abstract not provided.

Abstract not provided.

Abstract not provided.

Abstract not provided.

Abstract not provided.

Advancements in machine learning (ML) and deep learning (DL) have enabled imaging systems to perform complex classification tasks, opening numerous problem domains to solutions driven by high quality imagers coupled with algorithmic elements. However, current ML and DL methods for target classification typically rely upon algorithms applied to data measured by traditional imagers. This design paradigm fails to enable the ML and DL algorithms to influence the sensing device itself, and treats the optimization of the sensor and algorithm as separate sequential elements. Additionally, this current paradigm narrowly investigates traditional images, and therefore traditional imaging hardware, as the primary means of data collection. We investigate alternative architectures for computational imaging systems optimized for specific classification tasks, such as digit classification. This involves a holistic approach to the design of the system from the imaging hardware to algorithms. Techniques to find optimal compressive representations of training data are discussed, and most-useful object-space information is evaluated. Methods to translate task-specific compressed data representations into non-traditional computational imaging hardware are described, followed by simulations of such imaging devices coupled with algorithmic classification using ML and DL techniques. Our approach allows for inexpensive, efficient sensing systems. Reduced storage and bandwidth are achievable as well since data representations are compressed measurements which is especially important for high data volume systems.

Rigorous characterization of the performance and generalization ability of cyber defense systems is extremely difficult, making it hard to gauge uncertainty, and thus, confidence. This difficulty largely stems from a lack of labeled attack data that fully explores the potential adversarial space. Currently, performance of cyber defense systems is typically evaluated in a qualitative manner by manually inspecting the results of the system on live data and adjusting as needed. Additionally, machine learning has shown promise in deriving models that automatically learn indicators of compromise that are more robust than analyst-derived detectors. However, to generate these models, most algorithms require large amounts of labeled data (i.e., examples of attacks). Algorithms that do not require annotated data to derive models are similarly at a disadvantage, because labeled data is still necessary when evaluating performance. In this work, we explore the use of temporal generative models to learn cyber attack graph representations and automatically generate data for experimentation and evaluation. Training and evaluating cyber systems and machine learning models requires significant, annotated data, which is typically collected and labeled by hand for one-off experiments. Automatically generating such data helps derive/evaluate detection models and ensures reproducibility of results. Experimentally, we demonstrate the efficacy of generative sequence analysis techniques on learning the structure of attack graphs, based on a realistic example. These derived models can then be used to generate more data. Additionally, we provide a roadmap for future research efforts in this area.

Abstract not provided.

Abstract not provided.

Abstract not provided.

Abstract not provided.

Considerable effort is currently being spent designing neuromorphic hardware for addressing challenging problems in a variety of pattern-matching applications. These neuromorphic systems offer low power architectures with intrinsically parallel and simple spiking neuron processing elements. Unfortunately, these new hardware architectures have been largely developed without a clear justification for using spiking neurons to compute quantities for problems of interest. Specifically, the use of spiking for encoding information in time has not been explored theoretically with complexity analysis to examine the operating conditions under which neuromorphic computing provides a computational advantage (time, space, power, etc.) In this paper, we present and formally analyze the use of temporal coding in a neural-inspired algorithm for optimization-based computation in neural spiking architectures.

Abstract not provided.

Abstract not provided.