Özetçe-Çoklu hızlı uyarlamalı süzgeçleme, farklı hızlarda örneklenmiş iki veya daha fazla gürültülü gözlem verisinden yüksek çözünürlüklü işaretin geriçatılması problemi ile ilgilidir. Bu problemin çözümü için mevcut popüler bir yöntem, en küçük ortalama kareler (EKOK) yaklaşımına dayalı çoklu hızlı uyarlamalı süzgeç yapısını kullanır. Ancak yakınsama hızının yavaşlığı, bu yöntemin kullanımını kısıtlar. Bu çalışmada, düşük çözünürlüklü gürültülü gözlemlerden yüksek çözünürlüklü işareti elde etmek için EKOK yaklaşımına dayalı süzgeçleme yapısına alternatif çoklu hızlı normalize en küçük ortalama kareler (NEKOK) süzgeci önerilmiştir. Ses işareti ile gerçekleştirilen benzetim örneğinde, önerilen yöntemin var olan yönteme nazaran özellikle yakınsama hızı anlamında daha iyi sonuçlar ürettiği gözlenmiştir.
Anahtar Kelimeler -Çoklu hızlı uyarlamalı süzgeçleme; EKOK; NEKOK; uyarlamalı gürültü giderme.Abstract-Multirate adaptive filtering is related to the problem of reconstructing a high-resolution signal from two or more observations that are sampled at different rates. A popular existing method for solving this problem uses the multirate adaptive filter structure that is based on the least mean squares (LMS) approach. However, its low convergence rate restricts the use of this method. In this study, a multirate normalized LMS (NLMS) filter is proposed as an alternative to that of LMS based filter, for the reconstruction of the high-resolution signal from several low-resolution noisy observations. In the simulation example performed on an audio signal, it is observed that the proposed method leads to the better results than the existing method especially in the convergence rate.
High-resolution signal reconstruction from a set of its noisy low-resolution measurements is considered. As an alternative solution to this problem, a method employing the empirical mode decomposition (EMD) based denoising approach is proposed. In the framework of the proposed method, iterative EMD interval-thresholding based denoising procedure is applied to each noisy low-resolution measurement so as to filter the additive white Gaussian noise effect on it. We then synthesize the noise-reduced low-resolution signals to form the high-resolution signal. Unlike the method using the Wiener filter theory for highresolution signal reconstruction, the proposed method does not require knowledge of any correlation information about the desired high-resolution signal and its low-resolution versions. The validity of the proposed method is demonstrated by an audio signal reconstruction application.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.