Survey and Research Directions on Intrusion Detection in UNIX Environment

摘要

Although UNIX is considered a very stable and secure platform, the development of Intrusion Detection Systems is essential as current and future generations of hackers are continuously attempting to undermine its integrity. There are few intrusion detection systems in UNIX for detecting multiple threats in a distributed networking environment. Researchers have applied different statistical models that involve data fusion. The most common and popular approaches include Bayesian theory, Dempster Shafer Evidence Theory, Parametric and Non-Parametric techniques, and Markov Chain. With few exceptions, almost all these detection models cater only for single threat. Thus, there is a genuine need for research on multisensor data fusion model in intrusion detection systems that enhance its capability to detect multiple simultaneous threats, particularly in the UNIX environment. In this paper, I'll survey existing intrusion detection system s and detection models in the literature, followed by a discussion of my research directions on intrusion detection in UNIX environment.