RF-Based Noise Data for Randomness Applications with Artificial Intelligence

Title: RF Noise-Based Data

Description: This dataset comprises audio recordings of ultra-high-frequency ambient noise stored in the lossless waveform format (WAW). 

The recordings were sampled at a frequency sample rate of 2.048 MHz and then provided at a downsampled audio rate of 48 kHz for compatibility 

and practical usage. The total length of the dataset is 01:30:29, consisting of approximately 260 million data points.

Dataset Information:

- File Format: WAW (Lossless Waveform)

- Length: 01:30:29

- Size: Approximately 500 MB (size may vary based on file system format)

- Frequency Sample Rate: 2.048 MHz

- Audio Recorded Sample Rate: 48 kHz

- Total Data Points: ~260 million

Additional Notes:

- The dataset includes preliminary information to facilitate preprocessing and analysis.

- The preliminary data may contain each recording's metadata, annotations, or relevant information.

- The audio recordings are downsampled from the original frequency sample rate of 2.048 MHz to 48 kHz for compatibility and ease of use.

- The training data is provided with a sample rate of 1567 and a length of 200, totaling 4000 instances.

- Users are encouraged to refer to the accompanying documentation and related research paper for detailed information on dataset usage, preprocessing steps, and potential applications.