Weighted vote algorithm combination technique for anomaly based Smart Grid Intrusion Detection systems
Joris Lueckenga, Dominik Engel, Robert Green · 2016
Intrusion Detection systems (IDS) are a crucial and necessary aspect of the smart grid, particularly when considering the possible attack vectors and their consequences. While there are many different approaches on IDS for Smart Grid, the benefits of an anomaly detection technique is still in discussion, due to its capability of detecting zero-day attacks and misuse. This paper proposes a weighted vote classification approach and a general weight calculation function to improve the detection performances of anomaly IDS systems. Initial results show that a combination technique is able to improve classifier performance by several percent.