| With the rapid increasing impact of the mobile Internet on people’s daily life in the current world, accessing the Internet via mobile devices is showing an obvious trend, which brings a large amount of network traffic at the same time. The mobile Internet has been changing people’s behavior in recent years. The development of Internet services such as mobile social network services, mobile advertising, and e-commerce etc can provide a more convenient approach to people’s activities via network data. Actually, most of the network data is dominated by the Hypertext Transfer Protocol. Therefore, it is necessary to analyze the network traffic, especially HTTP traffic, from the mobile Internet in order to obtain a good understanding of the current operation status of the network and user behavior habit, and then optimize network and provide targeted network services to people.The first chapter describes the development of HTTP. The second chapter introduces message format and communication process of HTTP. The third chapter proposes a data acquisition system based on real-time traffic. First, we illustrate the pluggable packet parser briefly. The system compiles each protocol to a dynamic link library to call, so it can realize the extension of protocol parsing easily. Then the implementation of the module of HTTP service is described. For the module, HTTP packets are matched following a same5-tuple and the result will be outputted to a log. Furthermore, the system is monitored using statistics data to detect some abnormal condition. Then the idea of extracting HEAD of HTML packets is also discussed. The logs are analyzed in the forth chapter. Some conclusions can be drawn by analyzing matching success rate, HTTP traffic and the pattern of packets. Then we take a deep analysis of the performance of this system and provide some solutions to improve its performance. In the last chapter, this thesis gives a summary and outlook on signaling research in mobile network. |
PDF Full Text Request