Mel-Frequency Cepstral Coefficient-Enabled Analysis of Doppler Effect on Road Traffic Noise from Nairobi City


Authors: Ochungo Akech Elisha, Osano Simpson Nyambane, Gichaga John Francis

Abstract: This study investigated the Doppler effect on road traffic noise levels in an urban setting by deploying Mel-Frequency Cepstral Coefficient (MFCC) to characterize acoustic signatures across vehicle mix classes. Using a dataset of audio recordings from Nairobi city’s road network sampled from 42 sites, the study adjusted the equivalent continuous sound levels (Leq) data for the analysis of the Doppler effect based on vehicle pass-by speeds, and extracted MFCC features. This was followed by semantic matching, which aligned the sample site locations with captured audio signal data, hence facilitating accurate speed assignments. Statistical analyses and visualizations revealed variations in noise levels and MFCC features by vehicle class and location. The result of the analysis presents mathematical formulations, comprehensive Doppler effect characterization, and visualizations in the form of graphical diagrams. The study has contributed knowledge on urban noise-scape scenario building using computational methods. This has confirmed that MFCC-enabled analysis of audio signal data is a novel approach for the characterization of road traffic noise hotspots in urban environments. The same can be used to inform active noise control strategies for cities to improve the welfare of urban residents.

Pages: 1-15

DOI: 10.46300/91016.2026.10.1

International Journal of Neural Networks and Advanced Applications, E-ISSN: 2313-0563, Volume 10, 2026, Art. #1

PDF DOI XML

Certification