SIGNAL RECONSTRUCTION AND ADAPTIVE WAVELET-BASED DENOISING FOR SINUSOIDAL PARAMETER COMPARISON

Abstract

Accurate reconstruction and parameter estimation of sinusoidal signals in noisy environments remains a fundamental challenge in modern signal processing, particularly in applications such as communication systems, spectrum monitoring, and advanced signal analysis. The presence of noise significantly degrades the reliability of parameter extrac-tion, especially under low signal-to-noise ratio (SNR) conditions, necessitating robust and adaptive processing techniques.

This paper presents an adaptive wavelet-based noise suppression approach com-bined with signal reconstruction for precise comparison of sinusoidal signal parameters. The proposed method operates in the wavelet domain, enabling effective attenuation of noise components while preserving the intrinsic structure and characteristics of the underly-ing sinusoidal signal. Special emphasis is placed on maintaining the integrity of amplitude and frequency components, which are critical for accurate parameter estimation.

Within a unified simulation framework, both Discrete Wavelet Transform (DWT) and Undecimated Wavelet Transform (UWT) - based denoising techniques are implement-ed and systematically evaluated. Adaptive thresholding is applied to the wavelet coeffi-cients to optimize noise suppression performance under varying noise conditions. The re-constructed signals obtained through this process allow for reliable extraction and compara-tive analysis of key sinusoidal parameters.

Simulation results demonstrate that the UWT-based approach provides superior performance in terms of reconstruction accuracy and preservation of signal features, partic-ularly in low SNR scenarios. Compared to DWT, the UWT method exhibits reduced distor-tion and improved stability due to its translation-invariant properties.

The proposed framework offers a flexible and scalable solution for sinusoidal sig-nal reconstruction and parameter comparison, making it highly suitable for practical appli-cations in noisy signal environments. Furthermore, the approach demonstrates a strong po-tential for real-time implementation, enhancing robustness and reliability in dynamic and interference-prone operational environments.