Network Traffic Supervision System Based on Feature Distribution Analysis
Dexian Chang · Jisuanji gongcheng · 2009
Most existing network traffic supervision systems focus on the volume of traffic, and do not analyze information contained in these data. This paper utilizes entropy to capture the change of network traffic feature distribution, such as source IP address, target IP address, target port, and analyzes network traffic from the standpoint of feature distribution. It implements a supervision system by using this method. Experimental results show that this system has a higher detecting rate and a lower fault rate.