Architecture-oriented regular algorithms for discrete sine and cosine transforms

We propose fast algorithms for a class of trigonometric transforms. These algorithms have a unified structure and simple data exchange similar to the constant geometry Pease (1968) algorithm derived from the Cooley-Tukey (1965) FFT. We can easily extend many of the parallel FFT approaches for these algorithms. The idea of the method is to localize the nonregularities peculiar to fast algorithms for other trigonometric transforms into the nodes of the Cooley-Tukey FFT type computational graph. Only the basic operation in the nodes of the computational graph will be different for different transforms. Thus, a simple programmable processor element for executing the node functions can be the basis for parallel constructs