A lattice network based canonic realization of IIR digital filters

In this paper, a novel lattice network-based structure is derived for a canonic realization of recursive digital filters. For an Nth order filter, unlike the existing lattice structures the proposed structure has 2N free parameters that can be used to improve the structure properties. The relationship between the transfer function coefficients of the filters and the parameters of the proposed structure is established. The optimal structure problem is formulated in terms of minimizing the signal power ratio and solved within {0, 1} space such that the resultant is canonic, possessing only 2N + 1 multipliers, when the one-multiplier lattice is used as the elementary building blocks. To demonstrate the design procedure, an example is presented. It is shown that the proposed structure yields a much smaller signal power ratio than the traditional canonic structures and that it outperforms the classical tapped and injected numerator lattice structures in terms of reducing implementation complexity and minimizing the signal power ratio.