Construction of an algorithm in parallel for the Fast Fourier Transform

TL;DR

This paper presents the development and implementation of a parallel FFT algorithm on a computer cluster, demonstrating its application in astronomical image processing and filtering.

Contribution

It introduces a parallel FFT algorithm executed on a cluster of 2^n computers using MacMPI, applied to astronomical image filtering and reconstruction.

Findings

Successful implementation of parallel FFT on a MacOS cluster

Effective filtering and reconstruction of astronomical images

Demonstrated scalability and practical application in astronomy

Abstract

It has been designed,built and executed a code for the Fast Fourier Transform (FFT),compiled and executed in a cluster of 2^n computers under the operating system MacOS and using the routines MacMPI. As practical application,the code has been used to obtain the transformed from an astronomic imagen,to execute a filter on its and with a transformed inverse to recover the image with the variates given by the filter.The computers arrangement are installed in the Observatorio Astronomico National in Colombia under the name OAN Cluster and in this has been executed several applications.