基于“大数据+AI学习”的4G小区负荷压降方法研究  

Research on 4G cell load pressure drop method based on "big data + AI learning"

作　　者：罗铭 涂子龙 王超[1] LUO Ming;TU ZiLong;WANG Chao(Hubei Transmission Bureau of China Telecom Corporation Limited,Wuhan 430000,Hubei,China)

机构地区：[1]中国电信股份有限公司湖北传输局,湖北武汉430000

出　　处：《长江信息通信》2022年第10期187-190,共4页Changjiang Information & Communications

摘　　要：文章以提升用户感知为导向,以4G小区KPI负荷为依据,通过大数据手段自动输出高负荷小区清单,并基于4G小区参数设置、网络结构、用户分布及覆盖情况设计出一套参数自动优化算法,通过机器学习模式输出高负荷小区负荷压降参考方案,助力网优人员初步判断压降举措。同时引入AI技术,在有5G覆盖的场景下,利用UE位置的潮汐性特点,实现不同时间段内基站内部参数的自动校正,达到覆盖最优,最终实现4G小区的负荷压降,进而提升用户感知。Since this article is oriented to improve user perception, based on 4G cell KPI load, a list of high-load cells is automatically output through big data means, and a set of parameters is designed based on 4G cell-level load, network structure, user distribution and coverage. The automatic optimization algorithm outputs the reference plan for the load pressure drop of the ultra-busy cell through the machine learning mode, which helps the network optimization personnel to make a preliminary judgment on the pressure drop measures. At the same time, AI technology is introduced. In the scenario with 5G coverage, the tidal characteristics of the UE location are used to realize the automatic correction of the internal parameters of the base station in different time periods to achieve the optimal coverage, and finally realize the load pressure drop of the 4G cell, thereby increasing the number of users. perception.

关 键 词：负荷压降 用户感知 自动优化 潮汐性 

分 类 号：TP312[自动化与计算机技术—计算机软件与理论]