Copyrights to these papers may be held by the publishers. The download files are preprints. It is understood that all persons copying this information will adhere to the terms and constraints invoked by each author's copyright. These works may not be reposted without the explicit permission of the copyright holder.
Pinit Kumhom (PhD. thesis, Electrical and Computer Engineering, Drexel University, 2001, Also Tech. Report DU-MCS-01-01, Drexel University, 2001)
Design, Optimization, and Implementation of a Universal FFT Processor
Preprint (8.2 MB)
There exist Fast Fourier transform (FFT) algorithms, called dimensionless FFTs (L. Auslander, J. Johnson and R. Johnson, Dimensionless Fast Fourier Transform Method and Apparatus, Patent #US6003056, issued Dec. 14, 1999.), that work independent of dimension. These algorithms can be configured to compute different dimensional discrete Fourier transforms (DFTs) simply by relabeling the input data and by changing the values of the twiddle factors occurring in the butterfly operations. This observation allows the design of a universal FFT processor, which with minor reconfiguring, can compute one, two, and three dimensional DFTs. In this thesis a family of FFT processors, parameterized by the number of points, the dimension, the number of processors, and the internal dataflow is designed. Mathematical properties of the FFT are used systematically to simplify and optimize the processor design, and to explore different algorithms and design choices. Different dimensionless FFTs have different dataflows and consequently lead to different performance characteristics. A performance model is used to evaluate the different algorithmic choices and their resulting dataflow. Using the performance model, a search was conducted to find the optimal algorithm for the family of processors considered. The resulting algorithm and corresponding hardware design was implemented using FPGA.Keywords: IP cores for FPGA/ASIC, Discrete/fast Fourier transform