Implementation of adaptive sample rate Kwan-Martin notch filter using efficient realizations of reciprocal and squaring circuit

Adaptive algorithms often involve the use of divide and square operations. Rather than implementing generic dividers and multipliers in the adaptive section for the adaptive sample rate notch filter, the authors have developed prototype architectures. They present a technique of optimizing an implementation of the squarer. In the reciprocal circuit presented, they show how a numerator (scaling) can be easily incorporated into its design to achieve a general purpose divider. Both realizations have been demonstrated in an adaptive sample rate notch filter using Xilinx reconfigurable field programmable gate arrays (FPGA). A comparison is made between generic circuits and the proposed techniques. The results provide a measure of the advantages of the proposed approaches for use in adaptive signal processing