A cross-layer design for video applications based on QoE prediction

In this paper, we propose a cross-layer design scheme for optimizing resource allocation of video applications based on Quality of Experience (QoE) prediction. A novel mapping model between Peak Signal-to-Noise Ratio (PSNR) and Mean Opinion Score (MOS) based on hyperbolic tangent function is presented to reflect the relation between objective parameters and subjective perceived quality. On the basis of this model, we propose a QoE prediction function and utilize the Particle Swarm Optimization (PSO) method to solve the resource allocation problem. Simulation results show that our QoE-based cross-layer design scheme leads to significant improvements in terms of maximizing user-perceived quality as well as maintaining fairness among users.