This Article 
 Bibliographic References 
 Add to: 
Fault-Tolerant Meshes with Small Degree
May 2002 (vol. 51 no. 5)
pp. 553-560

In this paper, we study the design of fault-tolerant networks for arrays and meshes by adding redundant nodes and edges. For a target graph G (linear array or mesh in this paper), a graph G^{\prime} is called a k-fault-tolerant graph of G if when we remove any knodes from G^{\prime}, it still contains a subgraph isomorphic to G. The major quality measures for a fault-tolerant graph are the number of spare nodes it uses and the maximum degree it has. The degree is particularly important in practice as it poses constraints on the scalability of the system. In this paper, we aim at designing fault-tolerant graphs with both a small degree and a small number of spare nodes. The fault-tolerant graphs we obtain have degree O(1) for arrays and O(\log^3 k) for meshes. The number of spare nodes used are O(k\log^2 k) and O(k^2/\log k), respectively. Compared to the previous results, the number of spare nodes used in our construction has one fewer linear factor in k.

[1] M. Ajtai, N. Alon, J. Bruck, R. Cypher, C.T. Ho, M. Naor, and E. Szemerédi, “Fault Tolerant Graphs, Perfect Hash Functions and Disjoint Paths,” Proc. IEEE Symp. Foundations of Computer Science, pp. 693-702, 1992.
[2] N. Alon and F. Chung, “Explicit Construction of Linear Sized Tolerant Networks,” Discrete Math., vol. 72, pp. 15-19, 1988.
[3] Y. Aumann and M.A. Bender, “Fault Tolerant Data Structures,” Proc. IEEE Symp. Foundations of Computer Science, pp. 580-589, 1996.
[4] J. Bruck, R. Cypher, and C.-T. Ho, "Fault-Tolerant Meshes and Hypercubes with Minimal Numbers of Spares," IEEE Trans. Computers, vol. 42, no. 9, pp. 1,089-1,104, Sept. 1993.
[5] J. Bruck, R. Cypher, and C.T. Ho, "Wildcard Dimensions, Coding Theory, and Fault Tolerant Meshes and Hypercubes," IEEE Trans. Computes, vol. 44, no. 1, pp. 150-155, Jan. 1995.
[6] J. Bruck, R. Cypher, and C. Ho, “Fault-Tolerant Meshes with Small Degree,” SIAM J. Computing, vol. 26, no. 6, pp. 1764-1784, Dec. 1997.
[7] J. Bruck and C. Ho, “Fault-Tolerant Cube Graphs and Coding Theory,” IEEE Trans. Information Theory, Nov. 1996.
[8] P. Erdös, R. Graham, and E. Szemerédi, “On Sparse Graphs with Dense Long Paths,” Computers and Mathematics with Applications, vol. 1, 1975.
[9] J. Hastad,T. Leighton,, and M. Newman,“Fast computation using faulty hypercubes,” Proc. 21st ACM Symp. Theory of Computing, 1989.
[10] C. Kaklamanis, A.R. Karlin, F.T. Leighton, V. Milenkovic, P. Raghavan, S. Rao, C. Thomborson, and A. Tsantilas, “Asymptotically Tight Bounds for Computing with Faulty Arrays of Processors (extended abstract),” Proc. IEEE Symp. Foundations of Computer Science, pp. 285-296, 1990.
[11] F.T. Leighton,Introduction to Parallel Algorithms and Architectures: Arrays, Trees, Hypercubes.San Mateo, Calif.: Morgan Kaufmann, 1992.
[12] T. Leighton, B. Maggs, and R. Sitaraman, "On the Fault Tolerance of Some Popular Bounded-Degree Networks," Proc. 33rd Symp. Foundations of Computer Science, pp. 542-552, Oct. 1992.
[13] A.L. Rosenberg, “The Diogenes Approach to Testable Fault-Tolerant Arrays of Processors,” IEEE Trans. Computers, vol. 32, pp. 480-489, 1983.
[14] A.L. Rosenberg, “Fault-Tolerant Interconnection Networks: A Graph-Theoretic Approach,” Proc. Ninth Workshop Graph-Theoretic Concepts in Computer Science, pp. 286-297, 1983.
[15] H. Tamaki, “Construction of the Mesh and the Torus Tolerating a Large Number of Faults,” J. Computer and System Sciences, vol. 53, no. 3, pp. 371-379, 1996.
[16] H. Tamaki, “Efficient Self-Embedding of Butterfly Networks with Random Faults,” SIAM J. Computing, vol. 27, no. 3, pp. 614-636, 1998.

Index Terms:
Fault tolerance, interconnection networks, array, mesh, parallel computing
L. Zhang, "Fault-Tolerant Meshes with Small Degree," IEEE Transactions on Computers, vol. 51, no. 5, pp. 553-560, May 2002, doi:10.1109/TC.2002.1004594
Usage of this product signifies your acceptance of the Terms of Use.