An Objective Quality Assessment of Multimedia Service Based on Cross-Layer Optimal Mapping

Abstract

In order to provide the end-to-end service with QoS support, a novel objective quality assessment adopting the cross-layer optimized mapping from the key components' value of application layer to the subjective score of user-layer is presented in this paper. In this framework, the maximum likelihood estimation to the subjective score in subjective quality assessment with lower complexity is presented at first to avoid the complexity of the least-squares estimation. Then the optimal selection of the key component based on the highest loading is obtained, which take out the correlation among the application-layer parameters and simple the mapping form the key component to the application-layer equipment. At last, an optimal multiple regression line mapping based on the maximum likelihood decision is presented, which utilize the statistical quantifying outcomes of the subject score and adjust the coefficient of the mapping according to the fluctuation of those outcomes. This optimal mapping avoid the un-accuracy resulted from the ignorance of the statistics of the subjective score in the liner dynamic curve fitting. The objective quality assessment in this paper is facile to online monitor by the subject score or quality assess based on network scene. The computation and the simulation show that our suggestion outperforms the old one.