Performance analysis of a parallel Dantzig-Wolfe decomposition algorithm for linear programming

Abstract This paper employs the Dantzig-Wolfe decomposition principle to solve linear programming models in a parallel-computing environment. Adopting the queuing discipline, we showed that under very general conditions, the proposed algorithm speedup trends toward a limiting value as the number of processors increases.

[1]  Vaidy S. Sunderam,et al.  PVM: A Framework for Parallel Distributed Computing , 1990, Concurr. Pract. Exp..

[2]  Sylvie Gélinas,et al.  Locomotive assignment with heterogeneous consists at CN North America , 1997 .

[3]  Raisa E. Feldman,et al.  Predicting Speedup for Distributed Computing on a Token Ring Network , 1997, J. Parallel Distributed Comput..

[4]  Jack J. Dongarra,et al.  The PVM Concurrent Computing System: Evolution, Experiences, and Trends , 1994, Parallel Comput..

[5]  A. Sanghvi,et al.  Investment Planning for Hydro-Thermal Power System Expansion: Stochastic Programming Employing the Dantzig-Wolfe Decomposition Principle , 1986, IEEE Power Engineering Review.

[6]  Jacques Desrosiers,et al.  Crew Pairing at Air France , 1993 .

[7]  M. Aganagic,et al.  Security constrained economic dispatch using nonlinear Dantzig-Wolfe decomposition , 1997 .

[8]  Angappa Gunasekaran,et al.  Towards a portable and efficient environment for parallel computing , 1995 .

[9]  Gautam Mitra,et al.  Solving Large Scale Linear Programming Problems Using an Interior Point Method on a Massively Parallel Simd Computer , 1994, Parallel Algorithms Appl..

[10]  David Pisinger,et al.  An expanding-core algorithm for the exact 0-1 knapsack problem , 1995 .

[11]  John W. Chinneck,et al.  Computer Codes for the Analysis of Infeasible Linear Programs , 1996 .

[12]  Pasqua D'Ambra,et al.  Concurrent Banded Cholesky Factorization on Workstation Networks Using PVM , 1995, Parallel Comput..