Abstract | ||
---|---|---|
A parallel algorithm for solving complex hermitian Toeplitz linear systems is presented. The parallel algorithm exploits the special structure of Toeplitz matrices to obtain the solution in a quadratic asymptotical cost. Our parallel algorithm transfors the Toeplitz matrix into a Cauchy–like matrix. Working on a Cauchy–like system lets to work with real arithmetic. The parallel algorithm for the solution of a Cauchy–like matrix has a low amount of communication cost regarding other parallel algorithms that work directly on the Toeplitz system. We use a message–passing programming model. The experimental tests are obtained in a cluster of personal computers. |
Year | DOI | Venue |
---|---|---|
2006 | 10.1007/11758501_49 | International Conference on Computational Science (1) |
Keywords | Field | DocType |
low amount,parallel algorithm,communication cost,personal computer,experimental test,parallel solution,quadratic asymptotical cost,toeplitz linear system,complex hermitian,toeplitz matrix,toeplitz system,hermitian toeplitz linear system,message passing,programming model,linear system | Mathematical optimization,Algebra,Linear system,Computer science,Parallel algorithm,Matrix (mathematics),Personal computer,Toeplitz matrix,Hermitian matrix,Levinson recursion,Cauchy matrix | Conference |
Volume | ISSN | ISBN |
3991 | 0302-9743 | 3-540-34379-2 |
Citations | PageRank | References |
3 | 0.45 | 7 |
Authors | ||
3 |
Name | Order | Citations | PageRank |
---|---|---|---|
Pedro Alonso | 1 | 124 | 17.75 |
Miguel O. Bernabéu | 2 | 145 | 16.87 |
Antonio M. Vidal | 3 | 143 | 34.64 |