Mark Yampolskiy
University of South Alabama
Network
Latest external collaboration on country level. Dive into details by clicking on the dots.
Publication
Featured researches published by Mark Yampolskiy.
international conference on high confidence networked systems | 2013
Mark Yampolskiy; Péter Horváth; Xenofon D. Koutsoukos; Yuan Xue
The pervasiveness of Cyber-Physical Systems (CPS) in various aspects of the modern society grows rapidly. This makes CPS to increasingly attractive targets for various kinds of attacks. We consider cyber-security as an integral part of CPS security. Additionally, the necessity exists to investigate the CPS-specific aspects which are out of scope of cyber-security. Most importantly, attacks capable to cross the cyber-physical domain boundary should be analyzed. The vulnerability of CPS to such cross-domain attacks has been practically proven by numerous examples, e.g., by the currently most famous Stuxnet attack. In this paper, we propose taxonomy for description of attacks on CPS. The proposed taxonomy is capable of representing both conventional cyber-attacks as well as cross-domain attacks on CPS. Furthermore, based on the proposed taxonomy, we define the attack categorization. Several possible application areas of the proposed taxonomy are extensively discussed. Among others, it can be used to establish a knowledge base about attacks on CPS known in the literature. Furthermore, the proposed description structure will foster the quantitative and qualitative analysis of these attacks, both of which are necessarily to improve CPS security.
2012 5th International Symposium on Resilient Control Systems | 2012
Mark Yampolskiy; Péter Horváth; Xenofon D. Koutsoukos; Yuan Xue
Cyber-Physical Systems (CPSs) consist of as well as interact with cyber and physical elements. This creates multiple vectors for CPS-internal (i.e., within CPS) as well as for CPS-external (i.e., between CPS itself and its environment) Cyber-Physical Attacks. We argue that an effective Cyber-Physical Defense can only be elaborated if possible attacks on CPS can be identified and assessed in a systematic manner. In this paper, we focus on cyber-attacks only. Our contribution in this paper is the following. We assess the applicability of Data Flow Diagrams (DFD) for the systematic analysis of cyber-attacks against CPS. In this context, we introduce several extensions to DFD. We evaluate the analysis procedure by applying it on a comparatively simple example of a quad-rotor UAV. The selected UAV is fully functioning and contains multiple structural elements representative for more complex systems. At the same time, its simplicity enables an in-depth manual analysis. Our analysis shows that cyber-attacks executed against CPS can lead to various cyber-physical interactions. This, in turn, creates novel challenges for CPS defense. Finally, we outline the preliminary results of our work towards a Taxonomy of Cyber-Physical Attacks.
International Journal of Critical Infrastructure Protection | 2015
Mark Yampolskiy; Péter Horváth; Xenofon D. Koutsoukos; Yuan Xue
The security of cyber-physical systems is of paramount importance because of their pervasiveness in the critical infrastructure. Protecting cyber-physical systems greatly depends on a deep understanding of the possible attacks and their properties. The prerequisite for quantitative and qualitative analyses of attacks is a knowledge base containing attack descriptions. The structure of the attack descriptions is the indispensable foundation of the knowledge base.This paper introduces the Cyber-Physical Attack Description Language (CP-ADL), which lays a cornerstone for the structured description of attacks on cyber-physical systems. The core of the language is a taxonomy of attacks on cyber-physical systems. The taxonomy specifies the semantically distinct aspects of attacks on cyber-physical systems that should be described. CP-ADL extends the taxonomy with the means to describe relationships between semantically distinct aspects, despite the complex relationships that exist for attacks on cyber-physical systems. The language is capable of expressing relationships between attack descriptions, including the links between attack steps and the folding of attack details.
Proceedings of the 4th Program Protection and Reverse Engineering Workshop on | 2014
Mark Yampolskiy; Todd R. Andel; J. Todd McDonald; William Bradley Glisson; Alec Yasinsac
Additive Layer Manufacturing (ALM) is a new technology to produce 3D objects adding layer by layer. Agencies and companies like NASA, ESA, and SpaceX are exploring a broad range of application areas of ALM, which includes printing of device components, replacement parts, houses, and even food. They expect that this technology will greatly reduce production costs, manufacturing time, and necessary storage space. The broad variety of application areas and the high grade of computerization of this manufacturing process will inevitably make ALM an attractive target of various attacks. This research examines the problem of Intellectual Property (IP) protection in the case of outsourcing the ALM manufacturing process. We discuss the existing process and introduce a new model for the outsourcing of ALM-based production. For the proposed outsourcing model, focusing on IP protection, we present a risk assessment, specify requirements addressing mitigation of the identified risks, and outline approaches to implement the specified requirements. The fulfillment of the specified requirements will enable secure outsourcing of ALM production.
international conference on critical infrastructure protection | 2015
Mark Yampolskiy; Lena Schutzle; Uday K. Vaidya; Alec Yasinsac
Cyber-physical systems are under constant and increasing attacks as components of the critical infrastructure. Additive manufacturing systems are a new class of cyber-physical systems that produce threedimensional objects layer by layer. Agencies and companies such as NASA, the European Space Agency, General Electric and SpaceX have explored a broad range of application areas for additive manufacturing, including creating functional parts of safety-critical systems such as jet engines. The range of application areas and dependence on computerization makes additive manufacturing an attractive target for attackers.
automated software engineering | 2005
Jan Jürjens; Mark Yampolskiy
Designing and implementing cryptographic protocols is known to be difficult. A lot of research has been devoted to develop formal techniques to analyze abstract designs of cryptographic protocols. Less attention has been paid to the verification of implementation-relevant aspects of cryptographic protocols. This is an important challenge since it is non-trivial to securely implement secure designs, because a specification by its nature is more abstract than the corresponding implementation, and the additional information may introduce attacks not present on the design level. We propose an approach to determine security goals provided by a protocol implementation based on control flow graphs and automated theorem provers for first-order logic. More specifically, here we explain how to make use of assertions in the source code for a practical and efficient security analysis.
hawaii international conference on system sciences | 2017
Samuel Bennett Moore; William Bradley Glisson; Mark Yampolskiy
The utilization of 3D printing technology within the manufacturing process creates an environment that is potentially conducive to malicious activity. Previous research in 3D printing focused on attack vector identification and intellectual property protection. This research develops and implements malicious code using Printrbot’s branch of the open source Marlin 3D printer firmware. Implementations of the malicious code were activated based on a specified printer command sent from a desktop application. The malicious firmware successfully ignored incoming print commands for a printed 3D model, substituted malicious print commands for an alternate 3D model, and manipulated extruder feed rates. The research contribution is three-fold. First, this research provides an initial assessment of potential effects malicious firmware can have on a 3D printed object. Second, it documents a potential vulnerability that impacts 3D product output using 3D printer firmware. Third, it provides foundational grounding for future research in malicious 3D printing process activities.
2016 Resilience Week (RWS) | 2016
Samuel Bennett Moore; Phillip Armstrong; Todd McDonald; Mark Yampolskiy
3D printing is an emerging manufacturing technology that can produce objects with complex geometries. The aerospace, automotive, and healthcare industries increasingly use this technology to manufacture functional parts. Falling prices for desktop 3D printers are making this technology available to consumers. The widespread adoption of 3D printing technology means it will soon become the target of cyber-attacks. In this paper we analyze to what extent desktop 3D printers are vulnerable to such attacks. The software under scrutiny is open source 3D printer software that is broadly used with desktop 3D printers. Static analysis of the source code, dynamic analysis of the communication between 3D printer and computer, and architecture analysis reveal numerous vulnerabilities that can be exploited.
Sensors | 2015
Péter Horváth; Mark Yampolskiy; Xenofon D. Koutsoukos
In this paper, we present a system simulation framework for the design and performance evaluation of complex wireless cyber-physical systems. We describe the simulator architecture and the specific developments that are required to simulate cyber-physical systems relying on multi-channel, multihop mesh networks. We introduce realistic and efficient physical layer models and a system simulation methodology, which provides statistically significant performance evaluation results with low computational complexity. The capabilities of the proposed framework are illustrated in the example of WirelessHART, a centralized, real-time, multi-hop mesh network designed for industrial control and monitor applications.
Proceedings of the 10th Annual Cyber and Information Security Research Conference on | 2015
Paul Carsten; Todd R. Andel; Mark Yampolskiy; Jeffrey Todd McDonald
Vehicles made within the past years have gradually become more and more complex. As a result, the embedded computer systems that monitor and control these systems have also grown in size and complexity. Unfortunately, the technology that protects them from external attackers has not improved at a similar rate. In this paper we discuss the vulnerabilities of modern in-vehicle networks, focusing on the Controller Area Network (CAN) communications protocol as a primary attack vector. We discuss the vulnerabilities of CAN, the types of attacks that can be used against it, and some of the solutions that have been proposed to overcome these attacks.