Abstract:
We investigate techniques for minimising frequency estimator bias when using a Fourier-based frequency estimator that is applied to a time-varying sinusoidal signal. The ...Show MoreMetadata
Abstract:
We investigate techniques for minimising frequency estimator bias when using a Fourier-based frequency estimator that is applied to a time-varying sinusoidal signal. The frequency estimator that we use was developed by Quinn and was designed to operate on a constant tone signal. This motivates us to consider its performance when it is applied to a chirped signal, resulting in signal model mismatch. Two approximations to the so-called chirpogram approach to frequency and frequency rate estimation for a linear frequency modulated signal are proposed in this article. Of the two approximations, both are found to result in a computationally efficient technique that is suitable for the real-time applications that are envisaged in this work, but one technique is found to provide a much smaller bias in its frequency estimates.
Published in: 2019 13th International Conference on Signal Processing and Communication Systems (ICSPCS)
Date of Conference: 16-18 December 2019
Date Added to IEEE Xplore: 27 February 2020
ISBN Information: