基于用户偏好的网络质量QoE感知建模仿真分析

为了解决现有网络质量QoE感知模型数据粗差迭代次数多、线性回归参数小的问题,提出基于用户偏好的网络质量QoE感知建模仿真研究。依据用户偏好理论确定模型参数,并获取网络质量QoE感知数据,以此为基础,通过MCD算法判别并去除网络质量QoE数据粗差,以去除粗差的网络质量QoE数据为基础,利用ROI加权算法提取网络质量QoE数据特征,以得到的网络质量QoE数据特征为依据,将其代入多元线性回归方程计算网络质量QoE感知,实现了基于用户偏好的网络质量QoE感知。实验结果显示,与现有三种网络质量QoE感知模型相比较,构建的网络质量QoE感知模型降低了数据粗差迭代次数,提高了线性回归参数,充分说明构建的网络质量QoE感知模型具备更好的性能。

Modeling and Simulation Analysis of QoE Perception of Network Quality Based on User Preferences

In order to solve the problems that the existing network quality QoE perception model has more data gross error iterations and smaller linear regression parameters, a simulation study of network quality QoE perception model based on user preferences is proposed. Based on user preference theory, model parameters were determined and network quality QoE perception data were acquired. On this basis, gross errors of network quality QoE data were distinguished and removed by MCD algorithm. On the basis of removing gross errors of network quality QoE data, network quality QoE data features were extracted by ROI weighting algorithm to obtain network quality QoE data. Based on the data characteristics, the QoE perception of network quality wass calculated by substituting it into multiple linear regression equation, and the QoE perception of network quality based on user preferences was realized. The experimental results show that, compared with the three existing network quality QoE perception models, the constructed network quality QoE perception model reduces the number of data gross error iterations, improves the linear regression parameters, and fully demonstrates that the constructed network quality QoE perception model has better performance.