Digital Filters and Signal Processing: With MATLAB Exercises
Digital Filters and Signal Processing: With MATLAB Exercises
Computer-Based Exercises for Signal Processing Using MATLAB 5
Computer-Based Exercises for Signal Processing Using MATLAB 5
Digital Signal Processing with Field Programmable Gate Arrays with Cdrom
Digital Signal Processing with Field Programmable Gate Arrays with Cdrom
| Hi-index | 0.00 |
IIR digital filters can be implemented in a large number of different structures. In addition, there are various possibilities for the individual realization of these structures, depending on coefficient format, state variable format, quantization points and used overflow characteristics. An overview is given about the whole filter design procedure for cascades with second order sections (SOS) in the direct forms and in the state space form, including scaling and calculation of the noise figure for the selected structures and hardware realizations. For SOS in the state space form a noise optimized transformation is evaluated as a closed solution. The best pairing and ordering of poles and zeros within the cascade is calculated by "Dynamic Programming", an efficient optimization method for the allocation problem. Optimization is possible by the mean noise power as well as by the peak value of the noise power spectrum. Based on the exact solutions, rules of thumb are examined and further improved. The methods for optimized filter implementations are tested by an ensemble of filters with varying zero and pole positions. The results illustrate that, according to the transfer function, an appropriate choice of the SOS forms together with an optimization of pairing and ordering is useful for different hardware realizations.