Performance Optimization And Of Wireless Networks With Multi-Service

Recently, the rapid increasing of the number of mobile internet users has brought new demands for mobile communications. The previous target of enhancing transmission data rate of wireless networks could hardly provide various satisfying multimedia services. Quality of experience (QoE) extends the objective technical indicators to a new horizon of users’subjective perception of services. In the information age with fast spreading of mobile internets, operators, service providers and even equipment firms are focusing on the improvement of the QoE of their products to increase the market occupancy. Therefore, the key index in the wireless network is providing satisfying services with high QoE performance. Therefore, the exploration of QoE-oriented radio resource optimization is a significant issue for both research and application. As far, the research in this area is in a fledging period and needs more efforts and works to contribute to.The main concerned issue in this thesis is how to enhance the QoE performance in wireless networks with appropriate resource management and allocation approaches. The metric of QoE, i.e., mean opinion score (MOS), and the QoE modeling has been introduced. Two different kinds of technical factors in wireless networks, application QoS (Quality of Service) and network QoS, have been considered as the impacts on users’QoE. Then, the characteristics and QoE models of4popular services in the current mobile internet are analyzed, respectively.A QoE-oriented static resource management scheme has been proposed. The wireless network with multiple services is modeled with the consideration of both of the aspects of application and network levels. The proposed scheme is based on this model and can adapt the difference of the provided mobile services. Two objectives, i.e., maximum the sum of MOS and maximum the logarithm of MOS, are used for the optimization. In order to effectively reduce the complexity of optimization procedure, the dynamic programming has been adopted to solve the problem based on the model of multi-choice knapsack problem (MCKP). The obtained simulation results verify that the proposed scheme can significantly improve the QoE performance in multi-service wireless networks and guarantee fairness as well.Delay factor has been taken into account in this thesis in order to meet the demands of delay-sensitive services, such as web browsing. The web service is the main data service in the mobile network and thus has been used to analysis and design dynamic radio resource allocation algorithms. With the help of queuing theory, the wireless transmission system is modeled as a finite buffer system. Combining the advantages of both proportional fairness (PF) and time factor (TF), a novel radio resource allocation algorithm, i.e., PF-TF, has been proposed with the aim of shortening the end-to-end delay and improving QoE performance. The proposed algorithm has adjustable parameters to suit various channel conditions and can provide better QoE with higher MOS of web service. The gradient descent is adopted to find the optimal parameter setting, which can greatly reduce the computational complexity and thus improve the effectiveness of adaption procedure.