Abstract: A radix-16 fast Fourier transform (FFT) algorithm suitable for multiply-add instruction is proposed. The proposed radix-16 FFT algorithm requires fewer floating-point instructions than the ...