An Analysis on Performance of Data Fusion for Network Intrusion Detection System

Advances and Performance Analysis of Data Fusion Techniques for Network Intrusion Detection

Abstract

Fast advancement of networking technologies prompts an exponential development in the quantity of unapproved or malicious network activities. As a component of defense-in-depth, Network Intrusion Detection System (NIDS) has been relied upon to detect malicious practices. Right now, NIDSs are executed by different order techniques, however these techniques are not propelled enough to accurately detect unpredictable or manufactured attacks, particularly in the circumstance of confronting gigantic high-dimensional data. In addition, the innate imperfections of NIDSs, to be specific, high false caution rate and low detection rate, have not been viably illuminated. So as to take care of these issues, data fusion (DF) has been connected into network intrusion detection and has accomplished great outcomes. In any case, the writing still needs intensive analysis and assessment on data fusion techniques in the field of intrusion detection. In this manner, it is important to lead an exhaustive audit on them. In this article, we center around DF techniques for network intrusion detection and propose a particular definition to depict it. We audit the ongoing advances of DF techniques and propose a progression of criteria to look at their performance.