Robust scalable video multicast and performance issues in wireless LANs

Abstract: The successful deployment of multimedia streaming services, over a multiservice IP-based heterogeneous infrastructure, requires resource-efficient, scalable and reliable network transport. This can be achieved using end-to-end methods, e.g. adaptive application-level source and channel coding. Network assistance can also be employed, e.g. multicast routing and Quality of Service (QoS) provisioning mechanisms. Multicast transport in particular presents many challenges with respect to scalability, reliability and congestion control. Furthermore, with an increasing use of bandwidth-constrained and unreliable access networks such as Wireless LAN (WLAN) networks, implementing adaptive schemes which can utilize the scarce network resources is important. In this thesis, topics in the area of adaptive rate, congestion and error control schemes for video streaming applications over heterogeneous networks are studied. The contributions evaluate the end-to-end video quality gain from using a network-driven approach, based on router priority dropping and temporal video scalability. Furthermore, the end-to-end video quality gain from using a receiverdriven layered video multicast scheme is studied, using a dynamic Forward Error Correction (FEC) based error control scheme to enhance transport robustness. The scheme is evaluated in a WLAN infrastructure network setting. Finally, a model-based measurement study of WLAN access point forwarding behavior is included.