An artificial-neural-network-based QoE estimation model for Video streaming over wireless networks

In this paper, we present a no-reference, content-based Quality of Experience (QoE) estimation model for video streaming service over wireless networks. Since the impact of video quality impairments caused by both codec and network parameters is content-dependent, the cross-layer parameters, such as the bit rate, frame rate and resolution at the application layer, the packet loss rate at the network layer, video content features and the screen size of terminal equipment, are considered in the proposed QoE estimation model. Moreover, the video quality estimation model is based on radial basis function networks (RBFN) which is a feed-forward artificial neural network with excellent approximating ability. That is, the RBFN-based QoE estimation model is trained and tested with cross-layer parameters. Simulation results show that the RBFN-based QoE estimation model performs well in terms of high estimation accuracy, high Pearson correlation coefficient, low root mean square error, and small computational time.