A Deception Based Intelligent Intrusion Detection System for Detecting Threats of Exploits in Cloud Based Environments.

Abstract

Despite its numerous advantages, cloud computing faces major security threats with constantly evolving digital prints and attack-like patterns. Unfortunately, due to the share size and complexity of cloud computing, traditional approaches to Intrusion Detection Systems (IDS) have been shown to be rather defective in adapting to, identifying and mitigating threat in cloud based environment. While, anomaly-based IDS are plagued with misidentifying legitimate network activities or sometimes permitting sophisticated malicious traffic patterns, signature-based IDS on the other hand are less adaptive and practically ineffective against sophisticated attacks and advanced persistent threat (APT). This paper presents a unique design approach for deception-based intelligent Intrusion Detection Systems, which are better suited for operations in cloud based environments. Modelling and simulation was conducted using Application Characterization Engine and Flow Modelling Engine within OPNET modular to create runtimes of known attack types in a deception based environment. The machine learning scripts, attack codes and embedded socket and API integration scripts are presented in Python. The security framework was modelled with machine learning to further enhance its adaptability and predictive capabilities. Keywords: Cybersecurity, Intrusion Detection System, Deception techniques, Machine Learning