Application Based On Clustering Analysis In Intelligent Traffic Control System

Abstract:Facing the rapid growth of mobile data traffic, the intelligent pipeline network that has become a top issue for current is implemented in the operation. At present, PCC (Policy and Charging Control) technology is generally used in the industries, to limit the minority of users who are abusing network resources, to ensure that the majority of users for the rational use of limited network resources. Therefore, in order to realize effective traffic control, a set of traffic intelligent control strategy is needed to control mobile data service traffic from different user groups in busy areas and busy time. In this paper, a kind of custom methods about PCC traffic intelligent control strategy is proposed based on the theory of clustering analysis technology.The formulation of policy rules needs data platform to provide data support and this paper will use statistical data provided by signaling monitoring systemin PS domain as the basis of analysis. A refined process to formulatepolicy rules is proposed in this paper. Firstly use fuzzy K-Prototypes clustering algorithm for mobile customer segmentation. Then use hierarchical clustering algorithm based on fast grouping for classifying network equipments and analysis when is their busy and spare time. Finally combine with the segmentation of target users on PCC control and the result of clustering network equipments to formulate strategies about combined control on user in busy areas and busy time.The judgment and execution process of these rules is also proposed. Firstly determineuserswho try to connect to mobile network belong to which groups of users. Then judge whether the access area is busy and whether the current access time range in history busy time. Finally if these users try to access to mobile network in busy area and busy time, the strategy for specific user group will be taken effect. The method to develop rules which contains three factors among time, place and mobile user behavior, can improve the target user experience and ensure mobile network quality.Mobile user segmentation makes it possible to differentiated services. The analysisof busy and spare time about network equipments is helpful to realize localization of the control strategy. Combined policyrules proposed in this paper provide the feasible scheme andmake the intelligent control system work better.