Adaptive nonlinear filter using fractal geometry

Abstract

A new filter is described which uses fractal geometry in its synthesis. The important characteristic of the filter is that it allows some wanted high frequency parts of time series data to pass while rejecting other unwanted parts of the data.