Open AccessHigh‐order ridge reconstruction for more accurate signal estimate from time‐frequency representations
Author(s) -
Zhu Xiangxiang,
Zhang Zhuosheng,
Li Wenting,
Li Bei
Publication year - 2019
Publication title -
electronics letters
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.375
H-Index - 146
ISSN - 1350-911X
DOI - 10.1049/el.2019.0340
Subject(s) - signal reconstruction , signal (programming language) , ridge , algorithm , amplitude , window function , stability (learning theory) , computer science , iterative reconstruction , phase (matter) , mathematics , order (exchange) , signal processing , telecommunications , artificial intelligence , optics , physics , geology , paleontology , radar , spectral density , quantum mechanics , machine learning , programming language , finance , economics
Ridge reconstruction (RR) method is one of the most commonly used ways for non‐stationary signal reconstruction from time‐frequency representations. However, this method leads to a large reconstruction error when dealing with strongly amplitude‐modulated and frequency‐modulated (AM–FM) signals. To tackle this problem, in this Letter the authors propose a new and powerful reconstruction approach, named high‐order RR, by considering the high‐order derivative information of amplitude and phase of the signal. In particular, an explicit reconstruction formula for second‐order RR is derived. Simulation results demonstrate the superiority of the proposed approach in reconstruction accuracy when addressing AM–FM signals, as well as the stability for the selection of window width.