Fast Transformations and Kronecker Products
While on the subject of fast computational algorithms based on the Chinese Remainder Theorem and primitive roots (discussed in the preceding chapter), we will now take time out for a glance at another basic principle of fast computation: decomposition into direct or Kronecker products. We illustrate this by showing how to factor Hadamard and Fourier matrices — leading to a Fast Hadamard Transform (FHT) and the well-known Fast Fourier Transform (FFT).
Unable to display preview. Download preview PDF.