摘要
基于影响客户感知的主观KQI数据和客观网络KPI数据,建立多层次多维度的基于网格小区用户的动态评价体系。通过影响度基础因子、敏感因子、持续恶化因子等指标因子,区分场景和用户画像动态调整评估指标,应用熵值法计算各维度中指标的权重,引入变权函数动态调整初始权重,最后通过TOPSIS综合评估法,兼顾各维度对整体评价不同的影响程度,归一化后得出网格小区的综合评价得分。实际使用效果的统计数据验证表明,动态评价算法能够较好地反映网格小区用户的实际使用感知。
Based on subjective KQI data and objective network KPI data that affect customer perception,a multi-level and multi dimensional dynamic evaluation system based on grid cell users is established.Through such indicator factors as impact degree basis factor,sensitivity factor,and continuous deterioration factor,the evaluation indicators are dynamically adjusted to distinguish between scenarios and user portraits.The entropy method is applied to calculate the weight of indicators in each dimension,and a variable weight function is introduced to dynamically adjust the initial weight.Finally,the TOPSIS comprehensive evaluation method is used,taking into account the different impact degrees of each dimension on the overall evaluation,and the comprehensive evaluation score of the grid cell is obtained after normalization.The statistical data validation of the actual use effect shows that the dynamic evaluation al⁃gorithm can better reflect the actual use perception of grid cell users.
作者
傅玉明
刘德成
饶华阳
王政
FU Yuming;LIU Decheng;RAO Huayang;WANG Zheng(China Telecom Group Co.,Ltd.Sichuan Branch,Chengdu 610000,China;Ziyang Branch of Sichuan Telecom Co.,Ltd,Ziyang 641300,China)
出处
《通信与信息技术》
2024年第3期14-19,共6页
Communication & Information Technology
关键词
动态评价
熵值法
变权函数
TOPSIS综合评价
Dynamic evaluation
Entropy method
Variable weight function
TOPSIS comprehensive evaluation