Network User Action Detection based on PSD Signature through Encrypted Traffic Analysis

User action detection through network traffic analysis plays an important role in information security and network management. In previous studies, we proposed a rule-based user behavior detection system, and it showed high detection performance. However, due to the problem of high dependence on SNI information, detection performance was lowered in certain applications. This can significantly degrade detection performance when the SNI signature cannot be defined. In this paper, we additionally use the PSD signatures to solve the problem of high SNI dependence in previous studies. To verify the proposed method, a detection performance comparison experiment is performed for each signature, and the highest detection performance is shown when PSD is applied together.