PublicaHier finden Sie wissenschaftliche Publikationen aus den Fraunhofer-Instituten.
Simulation of digital filters with the aid of a universal program system
A matrix description of linear, time invariant, discrete time systems is used for the error estimation by simulation yielding a simulation algorithm which is independent of the individual filter structure to be simulated. Various quantization and overflow characteristics for fixed and floating point arithmetic in the dyadic number system with variable wordlengths can be utilized. Standard time domain analyses like limit-cycle-analysis, computation of frequency and impulse response and roundoff-noise-analysis for deterministic and random input signals can be performed. Examples from roundoff-noise- and limit-cycle-analyses demonstrate the applicability of the simulation system. From different filter structures realizing the same transfer function those filter structures exhibit a superior signal to noise ratio which are insensitive against coefficient quantization. A comparison between fixed and floating point arithmetic shows that floating point arithmetic possibly is to bepreferred if a prescribed signal to noise ratio is specified for a certain dynamic range.