Ventajas de la Transformada de Hartley en Problemas donde no Interviene la Transformada Rápida de Fourier

Abstract

The Hartley transform and its discrete version, both closely related to the Fourier transform, give real-valued results when applied to real data sets. Advantages of employing 1-D and 2-D Fast Hartley transforms instead of Fast Fourier transforms have been pointed out in the literature for certain cases. This paper shows that the Hartley transform is even more advantageous when a discrete Fourier transform needs to be estimated or calculated by methods other than the fast transformation, because of the less amount of storage required and the speed gained using real arithmetic. Two specific cases are presented: the estimation of a discrete Fourier transform by linear inversion, and the relation between frequency-domain continuation and singular value decomposition in potential field problems.