Virtual Machine Migration-Based Intrusion Detection System in Cloud Environment Using Deep Recurrent Neural Network

Abstract

Cloud system attracts users with the desired features, and in the meanwhile, cloud system may experience various security issues. An effective intrusion detection system is offered by the proposed Sail Fish Dolphin Optimization-based Deep Recurrent Neural Network (SFDO-based Deep RNN), which is utilized to identify anomalies in the cloud architecture. The developed SFDO is formed by integrating Sail Fish Optimizer (SFO) and Dolphin Echolocation (DE) algorithm. Virtual Machine (VM) migration and cloud data management are accomplished using ChicWhale algorithm. Some of the attribute features are collected from the cloud model such that these features are grouped using Fuzzy C-Means (FCM) clustering. The feature fusion process is carried out by a Deep RNN classifier that has been trained using the specified SFDO technique in order to achieve the intrusion detection mechanism. The approach with the lowest error value is thought to be the best approach for intruder detection, according to the fitness function. Using the BoT-IoT dataset, the proposed method’s accuracy, detection rate (DR), and false-positive rate (FPR) were assessed. The results showed that it outperformed the previous methods with values of 0.9614, 0.9648, and 0.0429, respectively.

Keywords:

• Cloud computing
• deep learning
• FCM clustering
• intrusion detection
• VM migration

Data Availability Statement

The data that support the findings of this study are openly available in BoT-IoT dataset at https://www.unsw.adfa.edu.au/unsw-canberra-cyber/cybersecurity/ADFA-NB15-Datasets/bot_iot.php.