基于AI算法的无线网络邻区关系优化研究与实践  

Study and Practice on Optimization of Wireless Network Neighborhood Relationship Based on AI Algorithm

作　　者：李张铮 陈锋[1] 董帝烺 Li Zhangzheng;Chen Feng;Dong Dilang(China Unicom Fujian Branch,Fuzhou 350000,China)

机构地区：[1]中国联通福建省分公司,福建福州350000

出　　处：《邮电设计技术》2021年第5期65-71,共7页Designing Techniques of Posts and Telecommunications

摘　　要：针对现有邻区优化方式的不足,基于现网数据引入XGBoost机器学习回归预测算法,通过学习具有自动邻区关系网络的两两小区切换占比建立预测模型,优化非自动邻区关系网络小区邻区关系。研究结果表明,基于AI算法的无线网络邻区关系优化能有效提高邻区优化效率,提升邻区关系的准确性。Neighborhood optimization is an indispensable and complex part of wireless network optimization,and traditional artificial optimization is a heavy burden on the operation of the operator's huge wireless base station.In view of the deficiency of the existing neighborhood optimization method,according to the existing network data,XGBoost machine learning regression prediction algorithm is introduced,and the neighborhood relationship of non-automatic neighborhood relationship network is optimized by learning the two-cell handover ratio of automatic neighborhood relationship network to establish the prediction model.The results show that the optimization of wireless network neighborhood based on AI algorithm can effectively improve the efficiency of neighborhood relationship optimization and improve the accuracy of neighborhood relationship.

关 键 词：XGBoost算法 小区切换占比预测 邻区优化 网络优化 

分 类 号：TN929.5[电子电信—通信与信息系统]