UWA-CP/Underwater-Acoustic-Channel-Repository

Underwater Acoustic Channel Repository

This Hugging Face dataset is a structured, checksum-preserving mirror of version 1.0 of the Underwater Acoustic Channel Repository. The original dataset was published by Zhengnan Li, Mandar Chitre, Diego Cuji, James Preisig, Andrew Singer, Milica Stojanovic, and Paul van Walree.

The collection contains measured underwater acoustic channel impulse responses (CIRs) from eight at-sea experimental groups. Channel and accompanying noise files remain in their original MATLAB representation. No train, validation, or test split is imposed: prediction windows, normalization, and evaluation splits belong to the downstream benchmark protocol rather than this source-data mirror.

Repository contents

Subset	Files	Size
Blue	21	3.840 GiB
Red	5	0.714 GiB
Yellow	10	3.929 GiB
Purple	20	7.961 GiB
Green	2	1.019 GiB
Black	2	0.151 GiB
Pink	6	4.131 GiB
Brown	1	0.615 GiB
Total	67	22.359 GiB

data/
├── index.csv
├── blue/
├── red/
├── yellow/
├── purple/
├── green/
├── black/
├── pink/
└── brown/

data/index.csv is the machine-readable catalog. Every source file has one row containing its subset, original filename, file role, repository path, byte size, MD5 checksum, and source DOI. The checksum and size values come from the immutable Zenodo record and can be used to verify the mirror.

Load the catalog

from datasets import load_dataset

catalog = load_dataset(
    "UWA-CP/Underwater-Acoustic-Channel-Repository",
    split="catalog",
)

Download a channel file

from huggingface_hub import hf_hub_download
from scipy.io import loadmat

path = hf_hub_download(
    repo_id="UWA-CP/Underwater-Acoustic-Channel-Repository",
    repo_type="dataset",
    filename="data/blue/blue_1.mat",
)
channel = loadmat(path)

The channel matrices are compressed complex-baseband CIRs. A time-varying phase vector stores the delay drift suppressed during compression. Reconstruction instructions and MATLAB/Python tooling are maintained by the original UWA Channels project.

Provenance and integrity

• Source record: Zenodo 21287414
• DOI: 10.5281/zenodo.21287414
• Source version: 1.0
• Source publication date: 2026-08-01
• Mirroring policy: original bytes and filenames are preserved; only directories and index.csv are added
• License: Creative Commons Attribution 4.0 International

This repository is a distribution mirror, not a replacement authority. The Zenodo record remains authoritative for provenance, citation, and release identity.

Citation

Please cite the original dataset:

@dataset{li_2026_uwa_channels,
  author    = {Zhengnan Li and Mandar Chitre and Diego Cuji and James Preisig and Andrew Singer and Milica Stojanovic and Paul van Walree},
  title     = {Underwater Acoustic Channel Repository},
  year      = {2026},
  version   = {1.0},
  publisher = {Zenodo},
  doi       = {10.5281/zenodo.21287414},
  url       = {https://doi.org/10.5281/zenodo.21287414}
}