Abstract. The General Packet Radio Service (GPRS) has become a standard to extend the services provided by the Global System for Mobile Communications (GSM). Through the introduction of packet switching technology, GPRS supports the use of packet data transmission, allowing a more efficient usage of the radio resources with a consequent improvement in the quality of service of subscribers. This paper presents some observations of the real operational GPRS network in a certain city by analyzing the characteristics of the data which was captured by monitoring system deployed in the Gn interface of GPRS core network. The distributions of the activated users and on-offline users were analyzed in this paper, as well as the online duration characteristic of mobile users. At last, the distribution of users and sessions was discussed, and finally some valuable conclusions were concluded for the mobile network operators.
IntroductionGPRS, based on packet switching technology, has been developed to enhance the GSM system. As the transition technology from GSM network to the 3rd generation mobile communication network, GPRS has significant advantages in many ways, such as high transmission speed and low cost [1].With wide spread use of smart mobile terminals, mobile network has become more and more important in people's life. The operators are also putting much more attention to the quality of network. It can be foreseen that the mobile internet traffic will increase explosively in the near future. So it is necessary to identify the mobile user's behavior and its impact on the network, which may help operators significantly in network maintenance and finally improve mobile subscriber's using experience.Analyses on the expected behavior of network have been performed in previous researches, which are essentially focus on measures like throughput or the transmission delay of a frame from the source to the destination. The results were normally established on simulation, for example, [2] presented experiments conducted at a GPRS testing platform, focusing on the performance evaluation of the GPRS signal operations. In [3] a GTP state machine was introduced, while the authors in [4] proposed a methodology for characterizing and identifying user behavior in online social networks.The initial idea of this paper was to capture packet-level traces from the GPRS network and use them to distil synthetic user behavior distribution. Our main contribution is that we present a sample set of findings and results from the exploration of real traces from the operational network with our monitoring system. And our insight based on large amount of data is valuable for network operators.The rest of the paper is organized as follow. At first, the deployment of system which is used to capture the data in GPRS network is introduced, and then the distribution of traffic generated by mobile users from GPRS network is analyzed and some user behavior characteristics are presented. Finally the conclusion is drawn in the last part of this paper.