Support for multimedia applications is a major objective of future high speed networks. Multimedia applications are usually resource intensive, have stringent quality of service (QoS) requirements, and in many cases involve large multicast groups. Multicasting enables these applications to scale to a large number of users without overloading the network and server resources. This paper focuses on developing low-cost, delay-bounded multicast trees to support the QoS requirements of multimedia applications. The approach taken in the development of the multicast trees is to decouple the cost optimization from bounding the delay by first building a low-cost tree and then handling any delay violations that may occur in the tree. Three new heuristics are proposed. The first two heuristics, delay-constrained low-cost inexpensive multicasting (SLIM) and SLIM+, use the least-cost path between the multicast nodes to incrementally build a multicast tree that satisfies the delay requirements of the multicast nodes. Their complexity is O(n3), where n is the number of nodes in the network. The third heuristic, K-SLIM, builds a set of k shortest paths and uses these paths in an attempt to further reduce the cost of the multicast tree without violating the delay requirements of the multicast nodes. Its time complexity is O(kn3 log(n)), where k, a user-defined parameter, denotes the number of shortest paths to be considered. Our simulation results show that K-SLIM on average outperforms other well-known heuristics. The results also show that SLIM+ produces low-cost, delay-bounded trees with an average cost close to the average cost of the trees produced by K-SLIM but with much lower processing overhead. Copyright 2001 Academic Press.
Document Type: Research Article
Department of Computer Science, University of Pittsburgh, Pittsburgh, Pennsylvania, 15260