A User-Centric Wireless Networks' Optimization Method for Data Service.

The optimization method based on traditional KPI statistics can't effectively meet the goal of improving user experience of quality (QoE). Based on QoE, this paper presents a kind of user-centric wireless networks optimization method that includes a users' QoE evaluation model and a locating the network problems method for data service. We propose a novel way of data analysis to build users' QoE evaluation model to allow operators to monitor users experience quality. Based on the technology of data fusion, we exploit a data mining algorithm for diging out the problems leading poor users' QoE and the corresponding probability from multisource data. Consequently, the operators can not only real-time monitor the entire network users' QoE and actively identify the wireless network problems, but also obtain reference guidance for the network optimization staff by utilizing the method. Obviously, this method ensures users' QoE and helps saving a large amount of manpower in networks optimization.