A Highly Parallel Algorithm For Computing The Singular Value Decomposition Using Optical Processing Techniques

Described in this paper is a highly-parallel algorithm (referred to as the Permutation Algorithm) for computing the singular value decomposition (SVD) of a matrix. This algorithm is an extension of an earlier algorithm developed by Brent and Luk. The Permutation Algorithm is unique in that it relies heavily on the repeated use of the more fundamental matrix-matrix multiply operation. Therefore, high-speed numerical optical processors currently being developed for performing the matrix-matrix multiply operation appear well suited for implementing the Permutation Algorithm as well.

[1]  J W Goodman,et al.  Fully parallel, high-speed incoherent optical method for performing discrete Fourier transforms. , 1978, Optics letters.

[2]  R P Bocker.  Matrix multiplication using incoherent optical techniques. , 1974, Applied optics.

[3]  B V Kumar.  Singular value decomposition using iterative optical processors. , 1983, Applied optics.

[4]  H. J. Caulfield,et al.  Rapid unbiased bipolar incoherent calculator cube. , 1983 .

[5]  I. Herstein,et al.  Topics in algebra , 1964 .

[6]  Harper J. Whitehouse,et al.  Parallel Processing Algorithms And Architectures For Real-Time Signal Processing , 1982, Optics & Photonics.

[7]  D Casasent,et al.  Iterative color-multiplexed, electro-optical processor. , 1979, Optics letters.

[8]  Franklin T. Luk,et al.  A Systolic Architecture for the Singular Value Decomposition , 1982 .

[9]  Peter S. Guilfoyle.  Systolic Acousto-Optic Binary Convolver , 1984 .

[10]  David S. Watkins,et al.  Understanding the $QR$ Algorithm , 1982 .

[11]  F. M. Hall.  An Introduction to Abstract Algebra , 1969 .

[12]  R A Athale,et al.  High accuracy matrix multiplication with outer product optical processor. , 1983, Applied optics.

[13]  K. Nichols,et al.  Spatial light modulation using electroabsorption in a GaAs charge‐coupled device , 1982 .

[14]  H. J. Caulfield,et al.  Optical Singular Value Decomposition For The Ax = b Problem , 1983, Other Conferences.

[15]  J.S. Harris,et al.  GaAs and related heterojunction charge-coupled devices , 1980, IEEE Transactions on Electron Devices.

[16]  R P Bocker,et al.  Electrooptical matrix multiplication using the twos complement arithmetic for improved accuracy. , 1983, Applied optics.