Home // CENICS 2020, The Thirteenth International Conference on Advances in Circuits, Electronics and Micro-electronics // View article

Predicting Noise Power in Gm-C Filters Through Machine Learning

Authors:
Malinka Ivanova

Keywords: machine learning; Gm-C filter; noise power; predictive model; signal flow graph

Abstract:
Noise level in Gm-C filters is connected to the dynamic range and to the ratio signal-noise. Noise depends on the design of the transconductor cell and filter topology. Predicting the noise power before filter realization could save engineers efforts and resources. The aim of the paper is to present a novel approach for predicting the total noise power in biquad low pass second order Gm-C filter through application of machine learning algorithms as data is taken from synthesized filter topology and filter mathematical model. Five machine learning algorithms: Artificial Neural Networks, Decision Tree, Random Forest, Gradient Boosted Trees, Support Vector Machine are applied for data training and they are evaluated in order to find the most suitable for this problem. The results show that the best solutions for solving this engineering task are Artificial Neuron Networks and Decision Tree algorithms, which are characterized with the best performance and high accuracy.

Pages: 1 to 7

Copyright: Copyright (c) IARIA, 2020

Publication date: November 21, 2020

Published in: conference

ISSN: 2308-426X

ISBN: 978-1-61208-823-5

Location: Valencia, Spain

Dates: from November 21, 2020 to November 25, 2020