An Adaptive Membership Algorithm for Application Layer Multicast

Abstract

Due to deployment difficulty of network layer multicast, application layer multicast is considered to be a good substitute for massive P2P video/audio streaming in large networks. However, in application layer multicast, the participating users join and leave the on-going session at will. Therefore, a scalable and reliable group membership management algorithm is necessary due to the highly dynamic nature of the overlay network, built on top of the Internet. Gossip-based algorithms seem to be a solution. However, most gossip-based membership management algorithms lack flexibility, and are unable to adapt to the everchanging network dynamics, imposing roughly the same amount of overhead on the network. A new adaptive gossip-based membership management algorithm is proposed to bridge the gap. This algorithm captures the changes of the network and adjusts the parameter settings dynamically, bringing adaptivity and reducing overhead. Simulation results indicate a maximum of 50% reduction can be achieved in terms of network overhead on core network components, such as backbone links and attached routers, without sacrificing reliability and scalability