Realization of adaptive digital filters using the Fermat number transform

An efficient realization of the block least mean-square adaptive digital filters using fast transforms is discussed. The computational complexities of the Fermat number transform realization and the fixed-point fast Fourier transform realization are compared and their convergence properties are studied through the computer simulation of practical applications.