Distributed resource allocation for network-supported FGS video streaming

In this paper, we propose a novel distributed resource allocation algorithm for fine granularity scalable video streaming. Our system model includes a proxy-based network infrastructure that is used for allocating locally the available bandwidth between the video rate and error control. Initially, we express analytically the expected decoder distortion as a function of both the source video streaming rate and the employed retransmission-based error control algorithm at each proxy. Next, we show that for this system configuration, the minimization of the decoder distortion is equivalent to a network flow control problem. We propose a distributed algorithm for solving this problem by applying Lagrange duality. The algorithm adapts the source streaming rate at the sender, while the number of retransmissions is optimized at each proxy individually. Simulation results demonstrate the effectiveness of the proposed scheme.