Computationally Efficient Finite Impulse Response Comb Filtering

摘要

A method and system that remove an unwanted signal and its harmonics from an input signal in a computationally efficient manner are disclosed. Embodiments include processing the FFT matrix to selectively zero-out rows of the matrix before multiplying the matrix with the Inverse FFT (IFFT) matrix. The resulting product (which is a sparse matrix) is then used to generate coefficients for a linear Finite Impulse Response (FIR) filter to process the input. The filtered output signal has the unwanted signal and its harmonics removed with minimal effect on a desired signal. The method produces a stable, physically realizable filter, requiring fewer computations than current methods.