Estimation of video QoE from MAC parameters in wireless network: A Random Neural Network approach

Abstract

Supporting real time video transmission over wireless networks is very challenging. The video flow characteristics can be vulnerable to the time-varying channel conditions and the unreliability of the MAC (Medium Access Control) can substantially degrade Quality of Experience (QoE) in wireless networks. QoE reflects the satisfaction level of a user towards a particular service or application. It can be a valuable input to service providers and help them to enhance the quality of the overall system. In this work, we present a methodology and a system based on Random Neural Network (RNN) to analyze the impact of different MAC-level parameters on video QoE over IEEE 802.11n wireless networks. At first, subjective tests are performed to correlate MAC-level parameters with user's perceived video QoE. Secondly, we propose a RNN technique to estimate the impact of these parameters on video QoE. The experimental results show that the proposed method can effectively measure the impact of MAC-level parameters on video QoE. The acceptable correlation between subjective QoE and estimated QoE verifies the validity of the obtained results.