Broadcasting free-viewpoint television over long-term evolution networks

Abstract

Recent advances in mobile device technology, cellular networks, and multimedia coding have paved the way for a wide range of innovative multimedia services that can be delivered to mobile users. An interesting application is free-viewpoint television (FTV), which allows the viewer to interactively select the scene viewing angle. Conveying this service over cellular networks is very challenging due to the large bandwidth requirements and limitations in computational power, memory, and battery resources on mobile terminals. Moreover, providing FTV over mo- bile systems necessitates new means for quality evaluation, which consider cellular network parameters. To address these challenges, we propose a scalable FTV broadcast architecture for long-term evolution cellular networks. To alleviate complex processing on mobile receivers, depth maps are estimated in the cloud and transmitted along the multiview video. Moreover, a novel depth image-based rendering technique is developed to improve quality while minimizing processing load on the mobile terminal, giving typical peak signal-to-noise ratio gains of 2 dB. Experimental results reveal that, for a satisfactory rendering quality, bit-rate savings of more than 70% are achieved compared with simulcast transmission. Furthermore, subjective tests indicate that this solution improves the perceived quality considerably. Finally, we study the impact of handovers induced in mobile systems on the quality of experience.