Quantifying Uncertainties of Security and QoS for Design of Power Grid Communications Systems

Reliable and secure communication, and high quality of service (QoS) (low latency, high availability, etc.) of data are critical to the successful operation of power systems. However, uncertainties, both aleatory and epistemic, are inherent to the system, which makes it difficult to secure the communication while fulfilling the QoS requirements. Nowadays, the administrators of the power grid's communication system have much flexibility in choosing a specific security scheme but the uncertainties make answering the question "what is the best scheme" or even "what is an adequate scheme" very difficult. Appropriately quantifying uncertainty and its effects on security is a key to choosing a sufficiently secure communication system for power grid while meeting QoS requirements. In this paper, we model aleatory and epistemic uncertainties using probability distributions and subjective logic respectively. Comprehensive quantification of the uncertainties can greatly sharpen administrators' understanding of the trade-offs between security and QoS and help them to select the most fitting security scheme to meet the QoS requirements. A case study demonstrates how our method works for two different application scenarios.