Datasets
Standard Dataset
Channel state information data of animal crossings on rural roads
- Citation Author(s):
- Submitted by:
- Samuel Ducca
- Last updated:
- Fri, 01/10/2025 - 13:44
- DOI:
- 10.21227/0t6t-8s40
- Data Format:
- Links:
- License:
13 Views- Categories:
- Keywords:
Abstract
This dataset is shared as part of the paper Towards scalable and low-cost WiFi sensing: preventing animal-vehicle collisions on rural roads, submitted to the IEEE Internet of Things Journal (IoT-J). It contains Wi-Fi Channel State Information (CSI) data from roadway crossings of small and large animals, persons and vehicles in rural environments.
To facilitate replication, we uploaded the same separated test and training datasets used for developing the machine learning models presented in our work. A previous version of the dataset used in preliminary work is linked in the dataset description. This new version contains twice the amount of samples and additional locations compared to the previous one.
Each sample from the dataset contains 500 frames of WiFi Channel State Information data, captured during a 5-second window. Each frame contains the amplitude information from the 52 WiFi subcarriers that transmit a modulated signal. This amounts to 26,000 features per sample.
We collected the CSI data using ESP32 boards, which were placed at a height of 70 cm and 12 meters apart from each other. To avoid bias towards a single environment, we collected data in four different locations, including paved and unpaved rural roads, a pasture and a gravel road.
The parquet files can be easily read and manipulated with python libraries such as pandas.
The data labels represent the following classes:
0 - Ambient background noise
1 - Person
2 - Car
3 - Dog
4 - Cow
