摘要
前台优化是无线网络优化中一个非常重要环节,但受限于现场信息获取困难、优化工具低效、人员水平参差不齐等因素,导致前台优化问题处理效率低、质量差;基于这些外场优化痛点问题,评估省内优化数据支撑平台,利用机器学习中的随机森林算法,对前期专项优化工作中总结的专家优化经验及问题案例进行训练,实现了一种对网络重要指标问题一键AI分析方法,并在APP上开发实现。该研究在全省全面推广应用,用于投诉处理、外场测试、质差小区处理等多个外场优化工作中,问题解决效率提升4倍,预计每月可节省外场优化项目成本325万元,实现网络降本增效的经营目标。
Foreground optimization is a very important link in wireless network optimization,but limited by such factors as difficulty in obtaining on-site information,low efficiency of optimization tools and uneven level of personnel,it leads to low efficiency and poor quality of foreground optimization problems;Based on these pain points of outfield optimization,the provincial optimization data support platform was evaluated,and the expert optimization experience and problem cases summarized in the previous special optimization work were trained by using the random forest algorithm in machine learning.A one key AI analysis method for important network indicators was realized and developed on the app.This research has been comprehensively popularized and applied in the whole province.It has been used in many field optimization work such as complaint handling,field testing,and poor quality community processing.The problem solving efficiency has been increased by 4 times.It is estimated that the cost of field optimization projects can be saved by 3.25 million yuan per month,and the business objective of network cost reduction and efficiency improvement can be realized.
作者
李柯
田华锋
Li Ke;Tian Huafeng(Hubei Network Excellence Center of China Mobile Communication Group Co.,Ltd Hubei Wuhan 430023)
出处
《长江信息通信》
2022年第12期219-221,共3页
Changjiang Information & Communications
关键词
随机森林
AI分析
专家经验
Random forest
AI analysis
expert optimization experience
