Abstract
Industry control systems (ICS) are considered as one of the inevitable systems in this contemporary smart world. In that supervisory control and data acquisition (SCADA) is the centralized system that control the entire grid. When a system is considered to be a whole and sole control, obviously an uncompromised security would be the prime. By having that as a major concern, a lot of research is being done on IDS security. In spite of that it has several cons including increased fake positive and fake negative rates, which will invariably lead to a larger chaos. To get rid of these problems, a weighted-intrusion based cuckoo search (WI-CS) and graded neural network (GNN) methods are proposed in this chapter. The key purpose of this chapter is to identify and categorize the anomalies in a SCADA system through data optimization. At initial stage, the collected real-time SCADA dataset is given as input. Then, by using the aforementioned proposed machine learning algorithms, these data are clustered and optimized. Later to find, the type of intrusion will remain as a further challenge and for that we propose HNA-AA algorithm. The investigational results estimate the efficiency of the system by considering sensitivity, false detection rate, precision, recall, Jaccard, accuracy, dice and specificity.
Shitharth Selvarajan
Lecturer in Cyber Security
My research interests include Cyber Security, Blockchain, Critical Infrastructure & Systems, Network Security & Ethical Hacking.