一种基于人工智能预测的个性化差小区优化方法研究  

Research on optimization method of poor cell based on AI

作　　者：李秀山 兰俊红 米凯 晓英 张勇 LI Xiu-shan;LAN Jun-hong;MI Kai;XIAO Ying;ZHANG Yong(China Mobile Group Inner Mongolia Co.,Ltd.,Huhehaote 010020,China)

机构地区：[1]中国移动通信集团内蒙古有限公司,呼和浩特010020

出　　处：《电信工程技术与标准化》2022年第S01期35-41,共7页Telecom Engineering Technics and Standardization

摘　　要：目前移动通信网络差小区的优化主要是通过人工分析各种特征指标,依据已有经验分析根因,给出相应的处理方案,均为事后分析。本文基于以预防为主的思路,利用人工智能预测算法库,采取竞争淘汰策略动态优选预测算法,根据预测值自动推理,定位出可能成为差小区的现无线网络小区,并输出预警信息和优化方法,使相关人员能够主动采取应对措施,防患于未然。经实际应用,该方法能够输出较为符合实际的优化建议,具有很高的推广价值。At present,the optimization of poor cells in the mobile communication network is mainly through the manual analysis of various characteristic indicators and the analysis of root causes based on existing experience,and the corresponding processing scheme is given.However,they are all ex post facto analysis and belong to mending the situation.This research is based on the idea of"treating the disease before it is ill",focusing on prevention,using the artificial intelligence prediction algorithm library,adopting the competitive elimination strategy to dynamically optimize the prediction algorithm,automatically inferring according to the prediction value,locating the existing network cells that may become poor cells,and outputting early warning information and optimization method research,so that relevant personnel can take measures actively to prevent problems before they occur.The practical application results show that this method can give more practical optimization suggestions,and has high promotion value.

关 键 词：规则库 决策树 推理引擎 预测 人工智能算法库A 

分 类 号：TN929.5[电子电信—通信与信息系统]