Description： This is the code of SPL paper "A Hybrid SFANC-FxNLMS Algorithm for Active Noise Control based on Deep Learning". You can find the paper at https://arxiv.org/pdf/2208.08082.pdf or at IEEE Xplore.

The paper proposes a hybrid SFANC-FxNLMS approach to overcome the adaptive algorithm's slow convergence and provide a better noise reduction level than the SFANC method. A lightweight one-dimensional convolutional neural network (1D CNN) is designed to automatically select the most suitable pre-trained control filter for each frame of the primary noise. Meanwhile, the FxNLMS algorithm continues to update the coefficients of the chosen pre-trained control filter at the sampling rate.

Platform: NVIDIA-SMI 466.47, Driver Version: 466.47, CUDA Version: 11.3

Environment: Jupyter Notebook 6.4.5, Python 3.9.7, Pytorch 1.10.1

Run Instructions:

To train the 1D CNN model, we generated 80,000 broadband noise tracks with various frequency bands, amplitudes, and background noise levels at random. Each track has a duration of 1 second. The synthetic noise dataset was subdivided into three subsets: 80,000 noise tracks for training, 2,000 noise tracks for validation, and 2,000 noise tracks for testing. The entire dataset is available at https://researchdata.ntu.edu.sg/dataset.xhtml?persistentId=doi:10.21979/N9/ETJWLU

If you don't want to train the model. The trained 1D model stored in "Trained models/model.pth" can be used directly.

Active noise control based on the proposed hybrid SFANC-FxNLMS algorithm on real-record noises. You can easily run "SFANC-FxNLMS for ANC.ipynb" The real noises are provided in "Real Noise Examples/"

Citation: If you find the hybrid SFANC-FxNLMS algorithm useful in your research, please consider citing: @ARTICLE{9761749, author={Luo, Zhengding and Shi, Dongyuan and Gan, Woon-Seng}, journal={IEEE Signal Processing Letters}, title={A Hybrid SFANC-FxNLMS Algorithm for Active Noise Control Based on Deep Learning}, year={2022}, volume={29}, pages={1102-1106}, doi={10.1109/LSP.2022.3169428}}

Contact Information: Zhengding Luo, Dongyuan Shi. The School of Electrical and Electronic Engineering, Nanyang Technological University, Singapore. (e-mail: [email protected]; [email protected])