Chapter 5 - Digital filter design

Abstract

This chapter reviews the design of FIR and IIR digital filters. The design of digital filters involves two distinct operations: (1) approximation and (2) realization. Approximation is used to derive a realizable system function from some idealized target specification. Realization concerns the formulation of a signal flow graph and its implementation in hardware or software. Neither of these operations necessarily has a unique outcome as a given target specification can usually be approximated by many system functions and a given system function can have many possible realizations. The differences between alternative realizations become particularly important when fixed point arithmetic is used and the effect of quantization becomes significant. The objective of most FIR digital filter design methods is to produce a causal and finite impulse response which can be realized by a transversal filter. Many design techniques for IIR digital filters have adopted ideas and terminology developed for analogue filters and is implemented by transforming the transfer functions of analogue prototype filters into the system functions of digital filters with similar characteristics.