Impact of the Market Infrastructure on the Security of Smart Grids

We study the resiliency of demand response (DR) systems to data integrity attacks. In particular, we compare two popular proposals for DR: 1) a centralized direct load control system, and 2) a decentralized dynamic prices system. Through an economic analysis, we show that decentralized dynamic pricing systems are more resilient to attacks that try to defraud (or damage) the system. On the other hand, we show that this resiliency comes with a disadvantage for detecting attacks, as distinguishing attacks from system failures is harder in systems with dynamic prices. We also propose penalties for the users who benefit from an attack. The penalties depend on the estimated losses of victims and can discourage rational attackers from launching attacks in the first place.