QoS/QoE Mapping and Adjustment Model in the Cloud-based Multimedia Infrastructure

Abstract

The quality of service (QoS) requirement and multicast service support from IP networks are two important factors for providing cloud-based multimedia services. However, QoS lacks an important element in characterizing multimedia services, namely, user perception. In this paper, we propose a QoS to quality of experience (QoE) mapping and adjustment model to translate the network QoS parameters into the user's QoE in the cloud-based multimedia infrastructure. The model is composed of three parts: QoE function, practical measurement and statistical analysis, and a simulated streaming video platform. We first discuss how to design the QoE function, and then use the practical measurement and statistical analysis to derive the optimum values of eight QoE parameters in the proposed QoE function. To map the network QoS parameters into the user's QoE, a simulated streaming video platform is used to denote a cloud-based multimedia infrastructure. Each multicast member that has guaranteed bandwidth in the simulated streaming video platform uses the QoE function to calculate its QoE score after watching the streaming video. If the QoE score is less than a derived lower bound value, it means that one of the multicast members has a low QoE. In this situation, the genetic algorithm is enabled to adjust the constructed diffserv-aware multicast tree to respond quickly to the degradation of QoE. The simulation results show that the user's QoE and network QoS are consistent with each other.