Broadcasting in a Hypercube when Some Calls Fail

Abstract In this paper, we consider broadcasting in the n -dimensional hypercube, under the assumption that a node can communicate simultaneously through all its ports. We prove that, if during each time unit at most n −1 calls fail, then the broadcasting time is n + o ( n ). even if the set of the faulty calls is different during each time unit.

[1]  Jhing-Fa Wang,et al.  Graph theoretic reliability analysis for the Boolean n cube networks , 1988 .

[2]  J. P. Hayes,et al.  Routing and broadcasting in faulty hypercube computers , 1988, C3P.

[3]  M. H. Schultz,et al.  Topological properties of hypercubes , 1988, IEEE Trans. Computers.

[4]  Quentin F. Stout,et al.  Hypercube message routing in the presence of faults , 1988, C3P.

[5]  R.D. McLeod,et al.  Percolation and Anomalous Transport as Tools in Analyzing Parallel Processing Interconnection Networks , 1990, J. Parallel Distributed Comput..

[6]  Pierre Fraigniaud,et al.  Fault-Tolerant Gossiping on Hypercube Multicomputers , 1991, EDMCC.

[7]  K. G. Shin,et al.  Message routing in an injured hypercube , 1988, C3P.

[8]  Parameswaran Ramanathan,et al.  Reliable Broadcast in Hypercube Multicomputers , 1988, IEEE Trans. Computers.

[9]  Krzysztof Diks,et al.  Optimal broadcasting in faulty hypercubes , 1991, [1991] Digest of Papers. Fault-Tolerant Computing: The Twenty-First International Symposium.

[10]  Chung-Chi Jim Li,et al.  Graceful Degradation on Hypercube Multiprocessors Using Data Redistribution , 1990, Proceedings of the Fifth Distributed Memory Computing Conference, 1990..

[11]  Robert Cypher.  Theoretical Aspects of VLSI Pin Limitations , 1993, SIAM J. Comput..