TD-LTE网络中大气波导干扰的分析与预测

发布时间:2018-07-17 作者:孙天宇,周婷,杨旸 阅读量:

[摘要] 使用来自于江苏移动的实时网络侧数据来分析大气波导干扰(ADI)的特征,同时结合网络侧数据与气象数据,使用两种机器学习方法对ADI强度进行预测,并相互比较。仿真结果表明:使用机器学习可以获得不错的ADI预测效果,当训练样本达到40 000条时,准确率与召回率分别可以达到72%与75%以上。

[关键词] 时分复用长期演进(TD-LTE);大气波导;机器学习;干扰预测

[Abstract] In this paper, the big data of network-side from the current operated network of China Mobile is used to analyze the characteristics of atmospheric duct interference (ADI). Combining network side data with meteorological data, two machine learning methods are used to predict the ADI intensity, and are compared with each other. The simulation results show that machine learning can achieve good ADI prediction effect. When the training sample reaches 40 000, the accuracy and recall rate can reach 72% and 75% respectively.

[Keywords] ime division-long term evolution (TD-LTE); atmospheric ducts; machine learning; interference prediction

下载阅览: PDF