Research on QoS Classification of Network Encrypted Traffic Behavior Based on Machine Learning

In recent years, privacy awareness is concerned due to many Internet services have chosen use encrypted agreements. order improve the quality of service (QoS), network traffic behaviors are classified based on machine learning discussed in this paper. However, traditional classification methods, such as IP/ASN (Autonomous System Number) analysis, Port-based and deep packet inspection, etc., can classify behavior, but cannot effectively handle traffic. Thus, paper proposed a hybrid (HTC) method combined with analysis inspection. Moreover, majority voting was also used quickly different QoS accurately. Experimental results show that HTC The accuracy be further improved by 10% K = 13. Especially when networking data using same protocol, differentiated code point (DSCP) mark.