A Heuristic Routing Algorithm for Degree-Constrained Minimum Overall Latency Application Layer Multicast

摘要

Application Layer Multicast (ALM) shifts multicast functionality from routers to end hosts and has the potential to address most problems associated with IP multicast. It has attracted wide attention in research community in recent years. However, as an end host based solution, the applicability of ALM to realtime applications such as streaming services is constrained by node bandwidth and transmission latency. How to guarantee QoS is still a challenge-able problem. In this paper, we think overall latency is a more effective metric for evaluating the QoS perceived by most users and explore the optimization problem of Degree-Constrained Minimum Overall Latency Spanning Tree (DCMOLST). We divide the optimization process into initialization phase and dynamic adjustment phase. In the former stage, we propose a heuristic algorithm through giving a more consideration to both transmission delay and node bandwidth, so as to avoid QoS degradation caused by single metrics. In the later, we present a set of distributed iterative optimizing operations for further optimization. Experimental results show that our proposal can improve overall performance efficiently and is able to cope with network dynamics.