A New Adaptive Filtering Algorithm for Signal Enhancement Problem with Hardware Implementation Using Arduino

Abstract

One of the most important issues in the area of digital signal processing is a signal noise cancellation that has been studied and given a considerable attention. In recent few years, adaptive filters have been used to cancel out different kinds of noise from analog signals based on many adaptive algorithms. The aim of this paper is to study and implement a real-time line enhancement problem (denoising) based on using Recursive Least Squares (RLS) and the modified version of RLS algorithms, which is proposed as a new adaptive algorithm (mRLS). Both are carried out in the microcontroller (Arduino). The main advantage of those algorithms is that they update the coefficient values of adaptive filters every single iteration until convergence is occurring. The performance analysis of the algorithms is assessed by showing the output of adaptive filters (denoised signals) and calculating the power signal to noise ratio (SNR). To verify their robustness, different values of random noise power are taken. It is observed and proved from the results, based on the above performance parameters, that the (mRLS) algorithm performs better than the (RLS) algorithm in terms of noise cancellation.