Using Self-Organizing Architectures to Mitigate the Impacts of Denial-of-Service Attacks on Voltage Control Schemes

Modern power systems are becoming increasingly decentralized, with a greater degree of observability provided through a network of sensors and local controllers in addition to existing centralized supervisory control and data acquisition platforms. However, the interconnectivity between sensors and controllers creates potential vulnerabilities which can be exploited by a cyber-attack. The majority of components installed on the grid were designed with little or no consideration for aspects of cyber-security and therefore leaving the network at risk of economic loss, asset damage or widespread blackouts. Present research in cyber-attack events and electrical grid resilience, often treats these in isolation. Furthermore, the ICT infrastructure in modern electrical networks is not tested as rigorously in terms of reliability and security as the physical assets. Therefore, an integrated approach is needed for the analysis of cyber-threats against power systems, linking the attack mechanisms in the ICT layer and the physical impacts at the electrical layer. This paper introduces a method of self-organizing communication architectures that for the first time has been applied to the problem of mitigating the negative impacts of denial of service cyber-attacks in the smart grid and demonstrates the benefits of this in a novel integrated environment connecting power system modeling and communication layer simulation. This paper demonstrates and quantifies the advantages of self-organization in terms of computational burden and voltage control in a distribution network experiencing multiple attack formats and increasing numbers of attackers.