EFFICIENT FAULT-TOLERANT MULTICAST SCHEME FOR HYPERCUBE MULTICOMPUTERS

This paper presents a fault-tolerant multicast scheme for hypercube mu lticomputers. The method is based on the routing capability informatio n that is stored in each node. In comparison with the previous schemes , this information is able to capture the fault status more precisely. Two multicast algorithms are presented in the paper. These algorithms multicast messages in an attempt to minimize derouting so that time o ptimality can be achieved. Moreover, the routing capability informatio n is used to guide derouting in an efficient manner when such needs ar ise. The amount of traffic incurred is addressed in the paper. The har dware design for the algorithms is also discussed. Extensive simulatio n has been conducted to evaluate the performance of the scheme. The re sults show the effectiveness of the proposed algorithms.