Scalable Video Coding in Fading Hybrid Satellite-Terrestrial Networks

Abstract

Broadband satellite multimedia (BSM) systems will be an integral part of the global information infrastructure as one of the major technologies providing both broadband access and broadcast services (Skinnemoen & Tork, 2002). Recent commercial deployments show that users not only would like to have access to value-added services (e.g., mobile internet, multimedia streaming, etc.) but are also willing to pay more for them and in particular for video services (Sattler). The introduction of video coding technology in the satellite application space opens up new and challenging topics; digital video applications have to face potentially harsher transmission environments than ones they were originally designed to work with (e.g., HDTV, Mobile TV), especially as regards traversing packet networks with the presence of satellite links. Towards approaching the satellite multimedia application delivery needs, H.264/MPEG4 Advanced Video Coding (AVC) (Ostermann et al, 2004), as the latest entry of international video coding standards, has demonstrated significantly improved coding efficiency, substantially enhanced error robustness, and increased flexibility and scope of applicability relative to its predecessors (Marpe et al, 2002). In the last decade, there is a growing research interest for the transmission and study of multimedia content over IP networks (Chou & van der Schaar, 2007) and wireless networks (Rupp, 2009). In an increasing number of applications, video is transmitted to and from satellite networks or portable wireless devices such as cellular phones, laptop computers connected to wireless local area networks (WLANs), and cameras in surveillance and environmental tracking systems. Wireless networks are heterogeneous in bandwidth, reliability, and receiver device characteristics. In (satellite) wireless channels, packets can be delayed (due to queuing, propagation, transmission, and processing delays), lost, or even discarded due to complexity/power limitations or display capabilities of the receiver (Katsaggelos et al, 2005). Hence, the experienced packet losses can be up to 10% or more, and the time allocated to the various users and the resulting goodput1 for multimedia bit stream transmission can also vary significantly in time (Zhai et al, 2005). This variability of wireless resources has considerable consequences for multimedia applications and often leads to unsatisfactory user experience due to the high bandwidths and to very stringent delay constraints. Fortunately, multimedia applications can cope with a certain amount of packet losses depending on the used sequence characteristics, compression schemes, and error concealment strategies available at the receiver (e.g., packet losses up to 5% or more