Title :
Adaptive wormhole routing in hypercube multicomputers
Author :
Lin, X. ; Esfahanian, A.-H. ; McKinley, P.K. ; Burago, A.
Author_Institution :
Dept. of Comput. Sci., Michigan State Univ., E. Lansing, MI, USA
Abstract :
We propose a uniform adaptive routing strategy for wormhole-routed hypercube networks that accommodates both unicast and multicast communication. Based on a node labeling method, the resultant routing algorithms are shown to be deadlock-free without requiring virtual channels. The order in which the destinations are visited is important to efficiency. We present an ordering algorithm, quadratic in the number of destinations, which is optimal in that it minimizes the traffic generated under the proposed paradigm. A greedy algorithm is also proposed for ordering the destinations, which requires less time and space to execute but creates more traffic than the optimal algorithm. Simulation results that evaluate the performance of the proposed routing algorithms are presented
Keywords :
hypercube networks; network routing; adaptive wormhole routing; deadlock-free algorithms; greedy algorithm; hypercube multicomputers; multicast communication; node labeling; ordering algorithm; paradigm; routing algorithms; traffic; unicast communication; uniform adaptive routing; virtual channels; wormhole-routed hypercube networks; Computer science; Greedy algorithms; Hypercubes; Intelligent networks; Labeling; Multicast algorithms; Routing; System recovery; US Department of Energy; Unicast;
Conference_Titel :
Parallel and Distributed Processing, 1993. Proceedings of the Fifth IEEE Symposium on
Conference_Location :
Dallas, TX
Print_ISBN :
0-8186-4222-X
DOI :
10.1109/SPDP.1993.395548