Survivable low-cost low-delay multicast trees

Abstract

The problem of finding survivable low-cost multicast trees having node-disjoint paths to destinations and satisfying delay requirements is an open area of research. Having two node-disjoint paths from the source to each destination makes a multicast tree resilient to node or edge failures. Satisfying delay requirements of the application ensures that the data delivered to the destination is fresh. Reducing the total cost of the multicast tree helps in lowering resource usage. We propose a new heuristic called dynamic weight optimal path pair (DW-OPP) for this problem. The heuristic constructs a node-disjoint low-cost multicast tree having low end-to-end delays. A user-specified parameter controls the performance of the heuristic. The parameter presents the user with a choice between lowering tree cost and achieving better delay performance. The performance of the new heuristic is evaluated empirically.