一种基于多数据源的LTE天馈隐患排查方法  

A Detection Method of Antenna Feeder Flaws Based on Multiple Data Sources

作　　者：石宙飞[1] 徐霖洲[1] 赵明伟[1] 

机构地区：[1]中国移动通信集团广东有限公司中山分公司,广东中山528403

出　　处：《移动通信》2017年第22期18-22,共5页Mobile Communications

摘　　要：针对目前依靠扫频数据发现LTE天馈隐患故障需配置大量人力物力的现象,创新性地提出一种通过对MR、MDT、第三方测试数据、网管告警、后台统计指标多种数据源进行分析,利用贝叶斯分类识别算法预估出天馈隐患故障概率的方法,可以节省测试成本,并且解决路测不能遍历全部小区的局限性。目前已经在多个地市验证了该方法的可行性和有效性。Since the use of sweep-frequency data to find out LTE antenna feeder flaws needs massive manpower and material resources, a detection method of antenna feeder flaws was innovatively proposed, in which multiple data sources including MR, MDT, third-party test data, network management alarm and background statistics are analyzed and the Bayesian classification recognition algorithm is adopted to estimate the fault probability of antenna feeder flaws. This method not only saves the testing cost, but also solves the limitation that the drive test cannot traverse the whole cell. At present, the feasibility and effectiveness of the method were verified in many cities.

关 键 词：多数据源 天馈隐患 LTE 贝叶斯估计 

分 类 号：TN929.53[电子电信—通信与信息系统]