Scalable Panoramic Wireless Video Streaming Relying on Optimal-Rate FEC-Coded Adaptive QAM
Yanqing Zhang,Jiankang Zhang,Yongkai Huo,Chao Xu,Mohammed El-Hajjar,Lajos Hanzo
DOI: https://doi.org/10.1109/tvt.2020.3008384
IF: 6.8
2020-01-01
IEEE Transactions on Vehicular Technology
Abstract:The demanding bitrate requirements of supporting flawless streaming of immersive video remains a challenge. To minimize the bandwidth without degrading the user's experience, we may opt for partitioning the panoramic video frame into numerous tiles and only transmit those covered by the predicted Field of View (FoV), but naturally, this philosophy critically hinges on the accuracy of the Head Movement Prediction (HMP). However, since the HMP is never 100% accurate, rebuffering of the missing portions due to misprediction is likely to introduce video freezes or artefacts, which may significantly degrade the users’ experience. Hence, instead of only transmitting the FoV tiles, Scalable Video Coding (SVC) comes to rescue. Explicitly, in SVC schemes, multiple layers having different importance provide a promising solution, where the basic quality of the entire panoramic video is supported by the Base Layer (BL) that only requires a low bitrate, while the Enhancement Layers (EL) are invoked for enhancing the quality of the predicted FoV. In this treatise, we propose coding rate adaptation assisted near-instantaneously Adaptive Quadrature Amplitude Modulation (AQAM) for layered panoramic video streaming. In our design, we categorize the video streaming into three priority classes according to the FoV and the SVC layer index, each of which is mapped to the most appropriate modulation mode determined by the instantaneous channel quality. Furthermore, we conceive an Evolutionary Algorithm (EA) assisted Forward Error Correction (FEC) coding rate optimization method for providing Unequal Error Protection (UEP) in order to maximize the uncoded source-rate according to the inter-frame and inter-layer decoding dependency for tile-based panoramic video streaming. The simulation results show that the proposed AQAM assisted UEP scheme configured by our EA assisted coding rate optimization algorithm significantly improves the overall video performance compared to its Equal Error Protection (EEP) counterpart, and provides perceptually pleasing video quality across a wide range of channel conditions by selecting the most appropriate modulation mode based on the instantaneous channel Signal-to-Noise Ratio (SNR).