Efficient content delivery scheme for layered video streaming in large-scale networks

Abstract

Providing efficient layered video streaming to heterogeneous users in varying network conditions requires dynamic bandwidth allocation, efficient data scheduling and incentives. In layered streaming, the video stream is composed of hierarchically encoded sub-stream layers namely the base layer and enhancements layers. We consider a scenario where receiver peers use a pull-based approach to adjust the video quality level to their terminal and network capacities by subscribing to a different number of layers. In this context, in order to take advantage of the available bandwidth in the network and to enhance end users Quality-of-Experience (QoE), we propose a novel approach that efficiently allocates sender peers׳ upload bandwidth to receiver peers. The upstream peer bandwidth is allocated depending on the quality level (requested layers) of the receiver peers, starting by allocating bandwidth for the lower layers first. In order to allocate bandwidth for a certain layer, an auction game is established to distribute the bandwidth among the receiver peers, where the sender peers “sell” their items (upload bandwidth) according to bids submitted by receiver peers. The main goal of this approach is to favor high priority peers while ensuring a minimum quality level to all peers. The proposed bandwidth allocation mechanism is paired with efficient scheduling mechanism for layered streaming. It aims to fully take advantage of the allocated bandwidth while respecting the layers dependency of the stream and the data blocks playback deadline. Extensive evaluations are conducted to compare our proposed algorithm with other bandwidth allocation strategies for layered video streaming. The obtained results show the effectiveness of our model in terms of video quality, useless chunks ratio and bandwidth utilization under different network/streaming conditions.