Abstract : Functional Description
The xFFT16 fast Fourier transform (FFT) Core computes a 16-point complex FFT. The input data is a vector of 16 complex values represented as 16-bit 2’s complement numbers – 16-bits for each of the real and imaginary component of a datum.
The 16-point transform engine employs a Cooley-Tukey radix-4 decimation-in-frequency (DIF) FFT [1] to compute the DFT of a complex sequence. In general, this algorithm requires the calculation of columns or ranks of radix-4 butterflies. These radix-4 butterflies are sometimes referred to as dragonflies. Each processing rank consists of 4 / N dragonflies. For N=16 there are 2 dragonfly ranks, with each rank comprising 4 dragonflies. The FFT processor input-data for the Core is a vector of 16 complex samples. The real and imaginary components of each sample are represented as 16-bit 2’s complement numbers. The data input and output buffers are stored internally. The phase factors used in the FFT calculation are generated within the Core. Like the input-data, the phase factors are kept to a precision of 16 bits.