CORDIC versus conventional logic for realisation of normalised lattice all-pass filters

In this paper an efficient pipeline architecture for normalised lattice all-pass filters is presented. The architecture is based on a special purpose CORDIC processing element which performs Given's rotations. A large number of filters is obtained exploiting its pipeline architecture and making use of the pipeline interleaving technique. The proposed architecture is compared with the conventional one, which uses multipliers/accumulators, and it is proved to be better regarding the hardware demands and the performance.