Cyber attack detection and mitigation process in cloud via deep hybrid model with selected feature set.

For ensuring the proper operation of CPS, it is required to obtain a solution to the security problem. Applications for cyber-physical systems (CPS) have a big impact on several industrial sectors. Growing in both quantity and variety are cyber attacks try to manage both the industrial process itself as well as data collection from CPS. Attacks on CPS must be identified and stopped in order to avoid financial loss, production disruptions, and possible threats to national security. This paper models a cyber-attack detection and mitigation framework using a hybrid classifier (CADFHC) including the steps like (1) Pre-processing, (2) Feature Extraction (3) Feature Selection (4) Detection & (5) Mitigation. The initial data is preprocessed via z-score normalization. Higher-order statistical features, mutual information, correlation, and raw features, will be retrieved in the feature extraction step. From the extracted feature set, an improved PCA will be used for selecting the appropriate features. In the detection phase, the hybrid model combining improved Bi-GRU & RNN is used for detecting the presence of an attack. After the attack detection process, the mitigation process will be performed by eliminating the attackers from the network. For this, modified distribution entropy function-based mitigation is performed. Finally, the effectiveness of the accepted model is compared with the conventional models with different metrics. The CADFHC approach has attained the most excellent outcomes (0.99) under the best case scenario. [ABSTRACT FROM AUTHOR]

Copyright of Multimedia Tools & Applications is the property of Springer Nature and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)