Search

Large scale power outage is typically the consequence of cascading failures propagated through a power system. To mitigate failure propagation, wide area monitoring and control system is introduced. However, as the physical system is tightly coupled with the cyber system, new threats are introduced due to possible cyber attacks and failures of the communication system.This paper presents a new framework for vulnerability analysis. Under this framework, we can identify the vulnerable components and the critical components of a cyber physical system. Distinct from previous work, our model considers the interaction between the different components of the cyber physical system, and models the dynamic evolving process of cascading failures. The impact of a component failure on the system is dynamically changing as the failure propagates. We formulate the vulnerability analysis problem as an optimization problem and use linear programming to solve it. Since instability is the reason of power outage, we use an instability index to measure the negative impact to the system. The results from this optimization problem suggest which components of the system are critical since their failure can most negatively impact the cyber-physical system.

Detection of sophisticated network scans, such as low and slow scans, requires correlation of large amounts of network data over long periods of time. The volume of data obfuscating such scans can be overwhelming and makes computation challenging. Such scans pose network security risks since identifying running services, the goal of executing such scans, is the first step in launching an attack on the scanned host. To detect sophisticated scans we propose the integration of graph feature extraction techniques with visualization to simultaneously optimize computational complexity and human analyst time. The integrated approach uses graph modeling and preprocessing to make visual displays easy to comprehend, and uses human intervention to avoid solving NP-hard computational problems while still providing real-time visualization.