Design and Implementation of DPI-Based User's Bebavior Perception System in LTE Network

Abstract

Based on telecom operators increasingly urgent demand for intelligent pipeline, a kind of implementing scheme of user behavior perception system with self-learning function was proposed, which analyzed the protocols of user plane on S1 interface in long term evolution（LTE）network. Aiming at the traditional monitoring system with poor user perception and insufficient statistical capacity, the proposed system achieved user's behavior refined analysis which combined protocol decoding and services call/transaction detail records（xDR）synthesis method with the advantages of current prevailing internet behavior analysis technologies—deep packet inspection（DPI）technique and focused crawler technology. With the verification of existing network data, this system not only achieves the goal, but also has the guiding value of meeting the demand for intelligent pipeline.

Keywords

• LTE network;S1 interface;deep packet inspection;focused crawler;user's behavior perception