Computing DFT using approximate fast Hartley transform

We propose an approximate fast Hartley transform (FHT) based method to compute the discrete Fourier transform (DFT) coefficients approximately. The approximate FHT is implemented using a periodic discrete wavelet transform (DWT). We find that the proposed method is computationally superior to both the radix 2 fast Fourier transform (FFT) and also the radix 2 approximate FFT algorithms.