Algoritmos de filtrado adaptativo basados en caracterización matemática para la reducción de ruido en señales de audio.

This research project focuses on how adaptive filtering algorithms can help reduce noise in audio signals. We are facing a significant problem, both globally and nationally: the decline in sound quality in communications, especially in places where there is a lot of interference. The project proposes to develop and simulate these algorithms, using mathematical principles to find solutions that maintain the clarity of the original signal. We follow a varied methodology: we conduct a literature review, carry out experiments, and apply what we have learned in real-world situations. To design, simulate, and evaluate algorithms, we use analytical, deductive, and experimental methods. To measure the performance of the filters, we use metrics such as Signal-to-Noise Ratio (SNR) and Mean Square Error (MSE). In this work, we describe how we implement and compare different adaptive algorithms, such as LMS (Least Mean Squares), RLS (Recursive Least Squares), and the Kalman filter. In addition, we highlight the importance of the Discrete Fourier Transform (DFT), which is an essential tool for analyzing signals in the frequency domain. The results indicate that the RLS filter and Adaptive Convolution are the most effective when it comes to removing noise and maintaining the fidelity of audio signals, establishing a solid basis for selecting the most appropriate filter according to the type of noise and audio quality requirements

• Hlavní autor: Rodríguez Gutiérrez, Kenneth Sebastián (author)
• Médium: bachelorThesis
• Jazyk: spa
• Vydáno: 2025
• Témata: Espectrograma; Fidelidad; Convergencia; Complejidad; Métricas
• On-line přístup: https://repositorio.uteq.edu.ec/handle/43000/9147