A Microprocessor Implementation of a 'Non-optimum' Filter
AP Clarke, Warren Marwood · 1980
A recursive 'non-optimum' filter has been implemented on multiple 16-bit microprocessors. The coefficients have been adjusted to be powers of two to replace multiplications with shift logic. The data is processed as 16-bit integers. A number of experiments indicated that the microprocessor system could satisfy requirements on data rate, passband ripple factors and out-of-bound attenuation normally only met with dedicated hardware and certainly not possible with a single processor implementation.