| With the popularity and development of Internet, Instant Messenger is used widely, and users are increasing quickly. On the Internet, Instant Messenger has become one of the largest business applications. In addition to easy for people to communicate, Instant Messenger has also brought security issues such as bad network behavior and business opportunities such as informing user behavior to provide special services. Through the identification, monitoring and statistical analysis, we can provide data about Instant Messenger to solve the problems.This paper first presents the most commonly used network packet identification techniques domestic and international in detail, analyzes the working principle of these methods, and compares with each other. Then it introduces the concept of network user behavior, characteristics and analysis methods. On this basis, the paper raises a network packet identification system. It explains the design and implementation process of original data collecting, software identification, application identification and save and analysis models. This system mainly uses DPI technology to identify Ali Want instant messenger software and a variety of specific application, monitors user behavior, and record the data for the analysis of user behavior. The models apply the hash table to improve the system of storage and search efficiency. In the end, the paper gets the experimental results of the system, and analyses the user data. |