Cell User Behavior Pattern Recognition Based on Code Stream Separation and Statistics
The reconnaissance and control of mobile communication cell user is a necessary means to ensure the safety of electromagnetic space and normal communication order.In view of the current situation that 5th generation(5G)mobile communication reconnaissance cannot be accurately ana-lyzed at the user level,a cell user behavior pattern recognition method based on code stream separa-tion and statistics is proposed.Through communication reconnaissance processing,not only the basic system signaling information can be obtained,but also user's specific signaling and identity information can be intercepted and analyzed.According to the identity and control information of the users,the upstream and downstream service code stream of the terminal are gained after de-modulating,descrambling and decoding.And the flow graph of the terminal can be further obtained from a statistical perspective.On this basis,the deep learning model is used to carry out transfer learning to identify its behavior as voice,short message or data.The behavior pattern recognition of cell users is statistically completed,and the recognition accuracy rate reaches 87.50％through ac-tual reconnaissance experiments validation,which lays an important foundation for the research in-to accurate reconnaissance and control of mobile communication user.
mobile communication reconnaissanceuser code stream separationbehavior pattern recognitiondeep learningtransfer learningflow graph
NETL
NSTL
万方数据
