QoE-based energy saving resource allocation for video streaming in wireless networks

Abstract

With the flourish of mobile video services, large amount of data requests from video subscribers and the necessity of assuring good user quality of experience (QoE) inevitably result in high energy consumption for network operators, giving rise to environmental as well as financial problems. The conflict between providing a satisfied user experience and reducing energy consumption makes it a tough job to bridge the gap between them. Nevertheless, the “marginal effect” between user experience and power consumption implies that a joint optimization of user experience and energy saving is possible and it will make great sense for network operators in order to maintain higher profits. In this paper, a joint optimization Resource Block (RB) and power allocation problem is formulated with three objectives, i.e., minimizing total power consumption, maximizing overall user experience and maximizing QoE fairness. To combat the contradiction of these objectives, Lexicographic method and Tchebycheff method are adopted in this paper. However, finding the optimal solution of the transformed single optimization problem is NP-hard due to the mixed combinatorial and nonconvex property. Therefore, an effective method is proposed, integrating evolutionary genetic algorithm (GA) and Lagrange dual method to find near optimal RB and power allocation solution. The effectiveness of the proposed scheme is validated by experiment simulations.