Enhancing Intrusion Detection System (IDS) by Using Honeybee Concepts and Framework

Abstract

Abstract-Intrusion Detection System has been studied for more than ten years. Though Artificial Intelligence (AI) techniques have been integrated to improve IDS but the success is still far from satisfaction. Thus, we believe a new strategy to improve IDS is badly needed. One of the solutions is by imitating the honeybee colony that can successfully protect their colony. In fact, the honeybee colony system and problems are quite similar with network system, and the way the bees protecting their colony can also be considered similar with the IDS in the network system. In this paper, we investigated the honeybee colony system as well as their detection system to get improvement methods for IDS engine in order to enhance IDS system for a better network defense. The adaptation of the honeybee protection and defense system itself is a new knowledge that can help other systems such as antivirus, antimalware, or even defense system to imitate the AI techniques in performing their functions. We train the proposed system with different types of attacks data and model different types of attack signatures. The performance of the proposed IDS is evaluated using NSL-KDD data set. The experiments show that the performance of the proposed model can detect novel intrusions and reduce false alarms.