Credit-based source-adaptive multilayered video multicast

Spatial and temporal variations in network bandwidth constraints constitute serious challenges to the multicast distribution of real-time video content. This paper presents a video multicast algorithm that combines multi-layered video encoding with feedback-based source adaptation in order to address varying bandwidth constraints in a multicast tree. The proposed algorithm uses a novel credit-based, hop-by-hop multicast ow control mechanism in conjunction with explicit rate congestion feedback. It achieves nearly optimal network utilization at the expense of packet losses that are isolated to low priority video content. The responsiveness, bandwidth utilization, scalability, video quality and fairness of the proposed mechanism are evaluated through simulations, and results suggest that the proposed mechanism is capable of providing a high quality video service in the presence of varying bandwidth constraints.