Subjective and objective evaluation and packet loss modeling for 3D video transmission over LTE networks

Abstract

The recent Long-Term Evolution (LTE) standard, thanks to the provision of high data rates, will enable bandwidth demanding multimedia applications like Three Dimensional (3D) video streams over wireless. In this paper, we study the transmission of 3D video sequences over LTE networks by modeling an LTE wireless network, which uses the best performing downlink packet scheduling strategy (Modified Largest Weighted Delay First (M-LWDF)) for delay sensitive applications, through the well-known Gilbert-Elliot Channel Model (GE). Also, we perform subjective and objective quality evaluation of the received 3D video sequences. The perceived video sequences are impaired with six packet loss rates produced by the corresponding LTE network. The three major contributions of the paper are: 1) GE parameters that represent real statistics of an LTE network 2) a subjective study on 3D video transmission over an LTE system modeled as a GE channel for different levels of impairments and 3) the provision of a publicly available database with 3D video sequences affected by packet losses distributed according to a GE model and associated Mean Opinion Score (MOS) values, to enable researchers to test their video quality evaluation algorithms.