CPS: Breakthrough: Collaborative: Securing Smart Grid by Understanding Communications Infrastructure Dependencies
pdf
Overview: The existence of complex interactions and interdependencies in cyber-physical critical infrastructures is well known, but a poorly understood phenomena. As an example, a smart grid (SG) is dependent on the cyber infrastructure for advanced metering, data collection/analytics and control, and this dependency brings in serious security and vulnerability implications. For example, an attacker who injects a malware in a telecommunication network may prevent a correct configuration of important components in energy systems, possibly leading to overloads and even a large scale blackout. Similarly, a contingency in the energy system, whether from natural events such as tree branches falling on a transmission line or from intentional attacks, may overload the telecommunication network, thus preventing a convergent communication and decision control loop, and potentially destabilizing the entire SG. This motivates us to understand, study and analyze in a holistic manner the vulnerabilities of SG system due to its intricate interactions with the communication networks. Specifically we will examine a variety of attacks, misconfiguration, malfunctioning, and failure scenarios in the telecommunications networks and characterize their impacts on the SG. The goal is to design a multi-level security framework that hardens the grid against malfunctions and attacks from the communications side and allows us to tolerate the events better. Our collaborative team includes interdisciplinary researchers from Temple University and Missouri University of Science and Technology with complimentary expertise in electrical (power) engineering, wireless communication, sensor networks, cybersecurity, and energy management.
Intellectual Merit: The proposal takes a holistic multilevel approach to understand and characterize the interdependencies between cyber-physical critical infrastructure, namely SG and communication systems, and securing them against attacks or failures. We start with the hardening of the application level SG communication protocols since the current protocols are notoriously easy to attack. Next, we consider robust state estimation techniques that exploit a steganography-based approach to detect bad data and compromised devices. In general, a SG attack may not only inject bad data, but also target more complex goals such as destabilizing the systems or generating large-scale blackouts. For this purpose, we propose trust-based attack detection strategies which combine the secure state estimation with power flow models and software attestation to detect and isolate compromised components. Finally, although compromised elements may be detected and removed, their effects may still generate failures that may propagate due to the interdependencies between the power grid and the supporting communication network. To this purpose, we propose reconfiguration strategies that combine light-weight prediction models, stochastic decision processes and intentional islanding to mitigate the spreading of failures and the loss of load. A unique aspect of SG security is the critical importance of timeliness, and thus a tradeoff between effectiveness of the mechanisms and the overhead introduced is crucial and will be studied via simulation and to a limited extent by actual experimentation.
Broader Impacts: The proposed research will not only help us understand how to design secure critical cyber-physical infrastructure, but also develop mathematical models, real-time data monitoring, management and analytic tools to assess situational awareness for intelligent decision control in the wake of security vulnerability of a SG coupled with communication systems. We will create a comprehensive plan for wide-scale dissemination and adoption of the developed techniques and tools to protect the SG. In addition to integrating the novel research findings into various courses taught at the collaborating institutions, a large number of students including female and under-represented minority, will be trained in SG security and interdependency analysis.
Submitted by sajaldas
on
Overview: The existence of complex interactions and interdependencies in cyber-physical critical infrastructures is well known, but a poorly understood phenomena. As an example, a smart grid (SG) is dependent on the cyber infrastructure for advanced metering, data collection/analytics and control, and this dependency brings in serious security and vulnerability implications. For example, an attacker who injects a malware in a telecommunication network may prevent a correct configuration of important components in energy systems, possibly leading to overloads and even a large scale blackout. Similarly, a contingency in the energy system, whether from natural events such as tree branches falling on a transmission line or from intentional attacks, may overload the telecommunication network, thus preventing a convergent communication and decision control loop, and potentially destabilizing the entire SG. This motivates us to understand, study and analyze in a holistic manner the vulnerabilities of SG system due to its intricate interactions with the communication networks. Specifically we will examine a variety of attacks, misconfiguration, malfunctioning, and failure scenarios in the telecommunications networks and characterize their impacts on the SG. The goal is to design a multi-level security framework that hardens the grid against malfunctions and attacks from the communications side and allows us to tolerate the events better. Our collaborative team includes interdisciplinary researchers from Temple University and Missouri University of Science and Technology with complimentary expertise in electrical (power) engineering, wireless communication, sensor networks, cybersecurity, and energy management.
Intellectual Merit: The proposal takes a holistic multilevel approach to understand and characterize the interdependencies between cyber-physical critical infrastructure, namely SG and communication systems, and securing them against attacks or failures. We start with the hardening of the application level SG communication protocols since the current protocols are notoriously easy to attack. Next, we consider robust state estimation techniques that exploit a steganography-based approach to detect bad data and compromised devices. In general, a SG attack may not only inject bad data, but also target more complex goals such as destabilizing the systems or generating large-scale blackouts. For this purpose, we propose trust-based attack detection strategies which combine the secure state estimation with power flow models and software attestation to detect and isolate compromised components. Finally, although compromised elements may be detected and removed, their effects may still generate failures that may propagate due to the interdependencies between the power grid and the supporting communication network. To this purpose, we propose reconfiguration strategies that combine light-weight prediction models, stochastic decision processes and intentional islanding to mitigate the spreading of failures and the loss of load. A unique aspect of SG security is the critical importance of timeliness, and thus a tradeoff between effectiveness of the mechanisms and the overhead introduced is crucial and will be studied via simulation and to a limited extent by actual experimentation.
Broader Impacts: The proposed research will not only help us understand how to design secure critical cyber-physical infrastructure, but also develop mathematical models, real-time data monitoring, management and analytic tools to assess situational awareness for intelligent decision control in the wake of security vulnerability of a SG coupled with communication systems. We will create a comprehensive plan for wide-scale dissemination and adoption of the developed techniques and tools to protect the SG. In addition to integrating the novel research findings into various courses taught at the collaborating institutions, a large number of students including female and under-represented minority, will be trained in SG security and interdependency analysis.