Priority-based Lateral Error Recovery Model In Application Layer Multicast

With the development of technology in ALM (Application Layer Multicast, ALM) and the continuous improvement for quality of service in network, error recovery is becoming more and more important, so an efficient error recovery model in ALM has become increasingly urgent.Firstly, this thesis reviews the classic error recovery technology which is not in ALM, secondly, some already existing error recovery models in ALM are summed up. In order to resolve the problem of high error correlation, implosion problem, and outage due to node-link failure in vertical error recovery, this thesis proposes a PLERM (Priority-based Lateral Error Recovery Model, PLERM). In PLERM, nodes are divided into a number of groups, each of which forms an independent group. As error correlation across groups is low, a node effectively recovers its error by"laterally"requesting retransmission from nearby nodes in other groups.Lastly, using Internet-like topology, it shows via simulation that PLERM is an effective error recovery model. As compared with traditional recovery models, PLERM achieves much lower average link stress and attains much lower residual loss rate.