COMPUTER SECURITY VULNERABILITY AS CONCERNS THE ELECTRICITY DISTRIBUTION GRID

摘要

The usual way of creating a cyber attack is through implementation of malware via the Internet. Among many types of malware, of special interest are those that enable eavesdropping on the activities within the computer, making it possible to define the software on which the computer is running. An adversary can benefit from this information in a way that is convenient for him or her. Here, we expose a new, entirely different way of eavesdropping and of monitoring the activities within the computer. It is based on measurement of the supply current taken from the electricity distribution grid. Because the computer, as many other electronic loads, is a nonlinear one, abundance of harmonics can be found in that current. Our discovery is the fact that the harmonic content is dependent on the type of activity within the computer, so, by proper analysis of the current waveform, one may recognize what is going on in it. We propose an artificial neural network-based method that unambiguously recognizes which software is running. We also propose a proper measurement procedure based on the technology we described in our earlier articles.