基于神经网络的MR精准定位  

MR Precise Location Based on Neural Network

作　　者：张进[1] 盛莉莉 冷俊[2] Zhang Jin;Sheng Lili;Leng Jun(China Unicom Jiangshu Branch,Nanjing 210029,China;China Unicom Nanjing Branch,Nanjing 210000,China)

机构地区：[1]中国联通江苏分公司,江苏南京210019 [2]中国联通南京分公司,江苏南京210000

出　　处：《邮电设计技术》2019年第11期16-19,共4页Designing Techniques of Posts and Telecommunications

摘　　要：4G业务和用户的发展,网络数据日趋增加,如何对网络中的海量数据,特别是MR和信令数据进行联合分析已经成为当前研究的热点。主要介绍中国联通江苏分公司通过依托MR大数据分析平台实现MR精确定位和定位结果应用方面的技术创新,主要包括神经网络指纹库算法以及全方位MR覆盖评估等方面,不仅大幅提高了MR定位精度,而且还提升了深度覆盖、业务价值、用户感知、竞对覆盖等全方面网络评估的可靠性。With the development of 4 G services and users,network data is increasing day by day. How to analyze the massive data in the network,especially MR and signaling data,has become a hot research topic. It mainly introduces the technical innovation of China Unicom Jiangsu branch in realizing MR precise positioning and application of positioning results based on MR Big data analysis platform,which mainly includes fingerprint algorithm and omnidirectional MR coverage evaluation. It not only greatly improves the accuracy of MR positioning,but also improves the reliability of all aspects of network assessment,such as deep coverage,business value,user perception,and competing coverage.

关 键 词：MR数据 定位方法 指纹数据库 神经网络定位 

分 类 号：TN929.5[电子电信—通信与信息系统]