Topology Attack Detection in Natural Gas Delivery Networks

Whenever a system operator is led to believe that the system topology is different from the actual system topology, we call this a topology attack. These attacks can cause tremendous damage from the result of improper network control. A number of algorithms to detect these attacks for electrical networks have been suggested in the literature, but such algorithms are lacking for gas networks. In this paper, we design a new detection algorithm to detect topology attacks in gas networks based on modifying the generalized likelihood ratio test in a novel way. Simulation results indicate the detection algorithm will provide good performance given a sufficient amount of data. Results also demonstrate that the performance also depends on the actual topology.