Research On Qoe-Oriented Wireless Resource Management Algorithms

With the rapid development of wireless communication systems and smart equipment, mobile internet services grows explosively, which brings in new demands for the design and implementation of communication networks. The previous target of enhancing Quality of Services (QoS) that is an objective indicator in wireless networks optimization could hardly convey the subjective experience of the end users to the current services, and may constrain the service capacity instead as a result of its strict QoS requirements. Quality of experience (QoE) expends the objective technical indicators in a traditional way into a brand-new horizon of users’subjective perception of services. QoE is defined by International Telecommunication Union as "a measure of the overall acceptability of an application or service, as perceived subjectively by the end-user." Moreover, as a uniform quality indicator in multi-services networks, QoE is able to unify many QoS indicators in a single service and also several kinds of different services impacting on user perception. Future wireless telecommunication systems should support more efficient resource utilization and higher users’satisfaction level. Radio resource management is an important approach to achieve this goal. Therefore, the exploration of QoE-based radio resource management is a significant issue for both research and application.According to the above demands of future wireless communication systems, this dissertation investigates on several radio resource management algorithms from the perspective of QoE including network selection, access control, resource management and dynamic resource scheduling joint. The main contents and contributions of this dissertation are presented as follows:Regarding with the heterogeneous characteristic of the future wireless communication networks, an energy efficient network selection multi-services resource allocation problem with group mobility is mainly discussed, and energy efficient resource allocation schemes in combination with adaptive sub-channels blackout are developed, including maximizing QoE algorithm and minimizing energy consumption algorithm with QoE guarantee. The two algorithms utilize sub-channels blackout to mitigate the inter-carrier interference in high mobility environment by turning off certain sub-channels when transmitting signals, so that channel quality is improved, a good balance can also be achieved between reduced transmit data rate and increased carrier to interference ratio, and energy can be saved linearly as a result of less active sub-channel. The proposed resource allocation algorithms consisting of inter-group resource allocation and inner-group sub-channels blackout. We also prove that under sub-channels blackout, the achieved throughput and perceptual QoE are quasi-concave in energy saving percentage, which can be figured out by greedy algorithm. Numerical results confirm the theoretical findings and demonstrate the promising energy-saving capability with satisfying QoE of the proposed resource allocation algorithms.Resource scheduling problem with maximizing QoE is first addressed. Then the proportional fair (PF) principle was adopted in QoE-based multi-services resource scheduling, termed as QoE-aware PF scheduling problem. Two integer programming problems are formed and then max-QoE scheduling algorithm and the QoE-aware PF scheduling algorithm are derived by solving the relaxed problems. Simulation results show that the proposed max-QoE can significantly enhance perceptive QoE but with unsatisfactory user fairness. QoE-aware PF scheduling algorithm performs well in terms of user QoE, and also increases the probability to be scheduled for cell-edge users, which results in a good balance between the overall QoE and fairness.