Microprocessor controlled development system for adaptive filtering using parallel processing and residue number arithmetic
Soderstrand, Cecilia Vernia · OSTI OAI (U.S. Department of Energy Office of Scientific and Technical Information) · 1981
The authors describe the realization of a fully programmable, microprocessor-controlled adaptive digital filter, which allows its user to investigate the behaviour of adaptive filters in real-time and at low cost. The user controls all filter parameters, such as filter structure, number of weights, adaptation step-size, and adaptation algorithm out of a bank of user defined algorithms. The filter performs parallel processing for high-speed adaptive filtering. Recent development of hardware for RNS arithmetic makes it feasible to realize the filter using a parallel multi-microprocessor structure, for large dynamic range, programmability, high-speed and low cost of processing. A new approach to the implementation of adaptation algorithms is proposed, and computer simulations show its capabilities to be quite promising. Applications of the new adaptive filter development system include standard non-recursive filtering, adaptive filtering, two-stage filtering, and totally adaptive filtering. The system also makes an excellent training aid for classroom or laboratory use. 21 references.