Qoe Based Performace Optimization And Design Of Wireless Network

With the rapid development of wireless communication systems and mobile Internet services, user demands for multimedia services are growing, optimization and design of wireless communication networks are also facing new challenges. All along, the main objective of wireless network optimization is to optimize the wireless network QOS (Quality of Service, QoS), ie some objective technical indicators, such as delay, jitter, throughput, and packet loss rate. However, these technical indicators can not be directly represented on users’subjective feelings. Therefore, ITU group proposed the concept of the user experience quality (Quality of Experience, QoE), which is defined as the user subjective satisfaction. QoE combines multiple QoS factors and can be integrated with a variety of user experience evaluation of multimedia services. The ultimate goal of the wireless network optimization is to provide a satisfactory user experience. Therefore, optimizing wireless network based on the user experience and design mechanisms, has important significance and practical value.The focus of the paper is that in the wireless cloud scenarios, how to improve system resource utilization and user service experience satisfaction through reasonable radio resource allocation and scheduling policies. This paper introduces the current status of wireless network optimization, and proposes the concept of user experience. Then, the paper describes the main factors affecting QoE (QoS objective factors and subjective factors users), quantitative methods of QoE (average score evaluation), QoE evaluation methods (subjective quality evaluation, objective quality evaluation and pseudo subjective quality evaluation), QoE evaluation model (sample data collection, data analysis and model validation). Meanwhile, the paper also analyse QoE optimization of wireless networks from four aspects including resource allocation, resource scheduling, network selection and access control.In this paper, the QoE-based radio resource scheduling scheme for delay-sensitive mobile cloud services is proposed. Firstly, the paper studies the main factors affecting QoS, and analyzes the QoS-QoE mapping model. The services will be modeled as an interactive mobile cloud services and a combination of channel and queue status of the system is adapted to achieve joint scheduling of uplink and downlink which greatly improves the user subjective experience. In the process of solving the Markov decision process-model, the system state space explosion. Therefore, in order to reduce the state-space system, the paper uses a stochastic learning algorithm. Finally, the simulation result show that the proposed algorithm improves QoE.This paper further explores how to improve network service experience of vehicular user satisfaction vehicular network. Firstly, a vehicular cloud network is proposed, in which vehicles are treated as both the service nodes and end users of cloud services. Then, the paper models resource scheduling process as a Semi-Markov decision process and get the optimal solution by solving the Bellman equation. Finally, compared with the traditional greedy algorithm, the proposed scheduling algorithm can effectively improve the average user experience.