Bio-inspired Metrics for Real-Time Detection of Denial-of-Service Power System Attacks
摘要
Cyber-attacks are occurring frequently as our power systems become more integrated with cyber components. This directly impacts the reliability of the power delivery capabilities of the physical grid. Protection strategies must encompass adaptable and robust quantitative assessment tools capable of keeping pace with technological advancements and emerging cyber threats while ensuring the continuous monitoring and maintenance of system health. This bio-inspired systems study offers unique insights into cyber-physical power systems disrupted by prevalent cyber network traffic pathways and Denial-of-ServiceService attacks. Ecological food webs, known for their robust disturbance response characteristics, are quantitatively analyzed using graphGraph and information theoryInformation Theory. Applying these characteristics to design and analyze physical power networks for resilienceResilience has proven successful when applied to information-flow cyber networks there are fundamental modelingModeling challenges. The proposed methodology enables cyber information flows to be compatible with the bio-inspired approach, helping to better understand time-dependent cyber-attacks within cyber-physical power system networks. The ecological metric known as the Degree of System Order, which has been directly linked to the resilienceResilience of ecological food webs, is calculated for cyber packet information and Round-Trip-Times. Disturbance scenarios highlight that each power system substation forms unique clusters based on their Degree of System Order at varying Round Trip Times. Moreover, by tracking packet transmissions per device, the approach uniquely enables the quantitative analysis of the location and spread of cyber-attacks across the system.