Digital filter for quadrature sub-sample delay estimation

In this paper we present a novel digital filter intended for quadrature sub-sample delay estimation based on the Hilbert (complex) filtering and fractional delaying of an incoming real-valued signal. It is a Hilbert Transform Filter (HTF) having variable fractional delay (VFD). The filter is based on a pair of rotated VFD filters in the Farrow structure. It is capable of performing the Hilbertian as well as VFD filtering of the incoming discrete-time signal at the same time. Thus one can substitute a hitherto used cascade of the HTF and the VFD filters with an aggregated HTF-VFD filter proposed here. The technique is simple to implement. The advantages lie in lower total delay introduced by the compound filter and in modular structure, which is composed of the same linear-phase rotated VFD Farrow subfilters. The overall rotated VFD filters in the pair differ only in the value of one parameter - the VFD. The proposed filter can be applied to adaptive quadrature sub-sample estimation of delay.