CYCLOSTIONARY FEATURE DETECTION USING WAVELET TRANSFORM

Abstract

This work presents the statistic cyclostationary feature detection algorithm that uses the analytic signal via Wavelet Transform (W.T.) and compares the proposed method with the cyclostationary feature detection algorithm using analytic signal via Hilbert Transform (H.T.). The method has the advantage of providing flexible sampling rate and step size of cyclic frequency. The wavelet transform based method presents higher signal-to noise ratio as compared to the Hilbert Transform method, as wavelet transform has time frequency localization characteristics.