Implementing branch-and-bound in a ring of processors
A set of personal computers is connected to form a ring structured parallel system: Each processor has access to its local memory and can exchange messages with its two ring neighbors.
A branch-and-bound procedure is implemented in Pascal to run in parallel on the ring and solve the Travelling-Salesman-Problem. Heuristics are developed to maintain a priority queue in a distributed heap. The computing times and speedups for 25 random graphs obtained with up to 16 ring members are discussed.
KeywordsTravel Salesman Problem Hamiltonian Cycle Global Memory Priority Queue Multiprocessor System
Unable to display preview. Download preview PDF.
- Finkel, R. and U. Manber, 1983, "DIB — A Distributed Implementation of Backtracking", Computer Science Technical Report #583, University of Wisconsin, MadisonGoogle Scholar
- Garey, M.R. and D.S. Johnson, 1979, "Computers and Intractability: A Guide to the Theory of NP-Completeness", Freeman, San Francisco, Calif.Google Scholar
- Lawler, E.-L. and D.E. Wood, 1966, "Branch-and-Bound Methods: A survey", Operations Research 14, pp. 699–719Google Scholar
- Li, G. and B.W. Wah, 1984, "Computational Efficiency of Parallel Approximate Branch-and-Bound Algorithms", Proc. of the 1984 International Conference on Parallel Processing, pp. 473–480Google Scholar
- Mohan, J. 1983, "A study in Parallel Computations: the Travelling Saelsman Problem", Technical Report CMU-CS-82-136(R), Dept. of Computer Science, Carnegie-Mellon University, PittsburghGoogle Scholar
- Monien, B., E. Speckenmeyer, O. Vornberger, 1986, "Superlinear Speedup for parallel Backtracking", submitted for publicationGoogle Scholar
- Monien, B., O. Vornberger, 1986, "The Ring Machine", submitted for publicationGoogle Scholar
- Vornberger, O., 1986, "Implementing Branch-and-Bound in a Ring of Processors", Technical Report Nr. 29, Dept. of Mathematics and Computer Science, University of Paderborn, W.-GermanyGoogle Scholar
- Wah, B.W. and Y.W. Eva Ma, 1984, "MANIP — A Multicomputer Architecture for Solving Combinatorial Extremum-Search Problems", IEEE Transactions on Computers, Vol. C-33, No. 5, pp. 377–390Google Scholar