基于无线信令指标的用户投诉定位方法

User Complaint Localization Based on Wireless Signaling Indicator

针对4G网络投诉数量骤增,用户投诉原因追溯难的问题,提出一种基于端到端网络业务质量评估模型的用户投诉原因精准定位方法。首先,针对用户投诉数据不平衡性的问题,在对LTE无线侧信令进行数据重构和稀疏特征提取的基础上,采用主动学习框架扩充用户感知数据库,为端到端网络业务质量评估模型提供数据基础;其次,采用BP神经网络关联用户投诉原因和无线侧信令指标,构建端到端网络业务质量评估模型;最后,基于端到端网络业务质量评估模型的结果,挖掘影响投诉原因的关键无线侧信令指标,获取关键无线侧信令指标的触发告警阈值,实现用户投诉原因的精准定位。实验证明,提出的方法具有一定的扩展性。

To deal with the rapid increase in the number of 4G network complaints and the difficulty in tracing the causes of user complaints,this paper proposes an accurate localization method of the causes of user complaints based on the end-to-end network service quality evaluation model.Firstly,for the imbalance issue of user complaint data,on the basis of data reconstruction and sparse feature extraction of LTE wireless side signaling,the active learning framework is used to expand the user perception database to provide data base for the evaluation model of the end-to-end network service quality.Secondly,BP neural network is used to associate the causes of user complaints with wireless side signaling indicators to establish the evaluation model of the end-to-end network service quality.Finally,based on the results of the end-to-end network service quality evaluation model,the key wireless side signaling indicators that affect the causes of complaints are excavate,and the triggering alarm threshold of the key wireless side signaling indicators is obtained to realize the precise positioning of the causes of user complaints.Experimental results show that the proposed method has a certain expansibility.