This paper studies the problem of constructing minimum-cost multicast trees with end-to-end delay and delay variation constraints to meet the quality of service (QoS) requirements of real-time interactive applications operated in high-speed packet-switched environments. The paper first discusses the routing problem with bounded delay along the paths from the source to each destination and bounded variation among the delays along these paths, and then presents a new algorithm based on Hopfield neural networks to optimize the multicast tree with delay and delay variation constraints. The simulations show that the proposed algorithm achieves its best performance in constructing constrained multicast tree in computer networks.
展开▼