Fast Fourier Transform - how to connect polynomial multiplication to signal processing

120 Views Asked by At

In this video, Reducible does a fantastic job explaining the derivation of the FFT algorithm and how it can be used to quickly multiply polynomials. More specifically, the input to the FFT algorithm is the coefficient representation of a single polynomial, and the output is its value representation at the nth roots of unity. Is there a way to connect this interpretation to signal processing? How does converting the coefficient representation of a polynomial to its value representation relate to converting samples of a signal in the time domain to the signal's frequency spectrum?