On the design of FIR digital data transmission filters with arbitrary magnitude specifications

A simple algorithm is presented for the design of FIR data transmission filters with arbitrary magnitude specifications. The technique involves the iterative solution of a set of linear equations as opposed to the more common approach of applying linear programming to the data transmission filter design problem. The design algorithm can accommodate conventional Nyquist filters or Nyquist filters with a non-negative frequency response that can be subsequently split into matched filter pairs using spectral factorization techniques. The filters can have equiripple stopbands or can be specified to satisfy arbitrary magnitude specifications, which is a common requirement in data transmission applications. Design examples are presented for filters satisfying FCC mask specifications for digital microwave radio channels