International Association of Engineers
In digital signal processing, the Discrete Fourier Transform (DFT) plays an important role in the analysis, design and implementation of discrete-time signal-processing algorithms and systems. The Fast Fourier Transforms (FFTs) are efficient algorithms to compute DFT. FFT is used widely in digital signal processing fields. Its performance is critical for many real-time applications. It is known that FFT algorithms have complexity of O (nlog2n), where n is the input size. Many new algorithms claim certain theoretical advantage; however, their real performance in application is questionable. The paper presents a systematic performance evaluation on different FFT software implementations.