Modern adversarial attacks are integrated with AI methods to target and evade network security solutions such as honeypots. Honeypots aim to learn from attackers behavior, however most of current honeypot systems are configured and managed statically wherein prior knowledge about the attackers is mandatory. Thus, intelligent honeypots have been designed to combat advanced human-based and AI-powered adversarial cyber attacks. This research aims to conduct a study on THREE (3) main objectives to define, design and evaluate the proposed solution. This research focuses to study current intelligent autonomic honeypot systems, define the concept of autonomic honeypot and introduce a new evaluation model with an application of reinforcement learning techniques. This helps to study and analyze different attacks with aim of estimating attackers next action.

Zoom link: https://vuw.zoom.us/my/ecspostgrad