基于信息熵的5G终端性能网格化评价算法

5G terminal performance gridding evaluation algorithm based on information entropy

5G终端性能评价在网络优化领域发挥着重要作用。本文提出了一种基于信息熵算法的网络大数据5G终端性能网格化评价方法。首先构建并改进了5G终端评价系统,将5G驻留时间作为KPI引入到终端评价系统中;其次提出了基于信息熵来训练评价系统因子权重。应用该方法,权重可以比基于终端评价系统的经验直接获得的结果更准确、更客观。同时在评价算法中引入了位置网格法,消除网络波动和用户分布的影响,最后采用终端抽样测试和投诉工单匹配的方法,验证算法的精准性。

Terminal performance evaluation plays an important role in the fi eld of network optimization. In this paper, a gridding evaluation algorithm based on network large data is proposed. Firstly, we constructed and improved the terminal evaluation system. Introduced the 5G dwell duration as KPI into the terminal evaluation system. Then the information entropy based algorithm is proposed to train the evaluation system factor weight. Applying this method, the factor weight can be more accurate and objectively than the result that directly obtained by experience. Based on the terminal evaluation system, we introduced location gridding method to the evaluating algorithm, which can remove the infl uence of network and user distribution. Finally, the terminal sampling test and complaint work order matching method are used to verify the accuracy of the calculation.