Attack detection and prevention in IoT-SCADA networks using NK-classifier

Supervisory control and data acquisition (SCADA) stands as a control system consisting of computers and networked data communications. At present, many industries use SCADA to monitor as well as control the processes. In recent days, numerous attacks are targeting these systems. Thus, the furtherance of high-security SCADA is much-needed one on account of its susceptibility to attacks centered on the architectural restriction. To identify these attacks, numerous classifications, optimization methods, and intrusion detecting systems (IDS) are posited. The chief drawbacks of this prevailing work are detecting accuracy, high training time, and security. For prevailing over these disadvantages, an NK-RNN classifier is proposed to recognize the intrusions in the SCADA method. Initially, the features from the datasets are organized, and the important attributes are chosen by utilizing the Elephant Herding Optimization (EHO). Secondly, the data, which is optimized, are grouped and classified by applying the NK-RNN classifier. Then, the outcomes, which are classified, are assessed and utilized to outcome prediction. In normal data, Caesar Ciphering is employed for the prevention of attacks and also the modified elliptic curve cryptography is employed for enhancing the security level. From the performance assessment, it is revealed that the NK-RNN method attains superior performance than the prevailing classification method along with IDS algorithms.