Datasets
Open Access
CSI Dataset towards 5G NR High-Precision Positioning
- Citation Author(s):
-
KaixuanGao
Harbin Engineering University: Harbin, Heilongjiang, CN - Submitted by:
- Kaixuan Gao
- Last updated:
- Fri, 06/16/2023 - 20:03
- DOI:
- 10.21227/jsat-pb50
- Data Format:
- Research Article Link:
- License:
5442 Views- Categories:
- Keywords:
Abstract
This is a CSI dataset towards 5G NR high-precision positioning,
which is fine-grained, general-purpose and 3GPP R18 standards complied.
The corresponding paper is published here (https://doi.org/10.1109/jsac.2022.3157397).
5G NR is normally considered to as a new paradigm change in integrated sensing and communication (ISAC).
Possessing the advantages of wide-range coverage and indoor-outdoor integration, 5G NR hence becomes a promising way for high-precision positioning in indoor and urban-canyon environments.
However, 5G Location studies are facing great obstacles due to the lack of commercialized 5G ISAC base stations that support positioning functions, as well as publicly available datasets.
To overcome this dataset deficiency, we make our 5G NR POSITIONING DATASET publicly available.
This dataset can be used for indoor positioning, indoor-outdoor-integrated positioning, NLoS, 5G channel estimation and other types of research, providing researchers with CSI-level position-related feature data.
If you'd like to learn more about our dataset, spend some time going through our paper for the model overview, generation method, 5G NR reference signal and many other subjects.
Also, we set up an open system for researchers to upload their own scene maps to obtain customized data sets.
Contact Us kaixuangao@foxmail.com (primary), gkx@hrbeu.edu.cn
keywords: integrated sensing and communication, ISAC, 5G, New Radio, 5G NR, massive MIMO, indoor Localization, indoor positioning, 5G positioning, 5G localization, CSI, channel statement information, ray-tracing, ray tracing, Machine Learning, Deep Learning, CNN, DNN, mmWave, sub 6GHZ, 3GPP
The dataset_[SNR]_[Scenario]_[date_time].mat contains:
1) a 4-D matrix, features, representing the feature data, and
2) a structure array, labels, labeling the ground truth of UE positions.
[SNR] is the noise level of features, [date] and [time] tell us when the dataset was generated.
The labels is a structure array. labels.position records the three-dimensional coordinates of UE (meters).
The features is a matrix, Ns-by-Nc-by-Ng-by-Nu, where Ns is the number of samples, Nc is the number of MIMO channels, Ng is the number of gNBs and the Nu is the number of UEs.
The value of Ng corresponds to the number of UEs in labels.
Release note
2023-05-15 :
1) Publish dataset files for Scenario 2 outdoor urban canyon.
2) Publish map files for Scenario 2 outdoor urban canyon.
3) Provide a data import example for Python users. (No need to convert .mat files to .csv files for direct use in Py)
2021-07-23 :
1) Recruit participants for a closed beta test.
2021-07-22 :
1)Expend our dataset with more CSI data with low SNR levels noise.
2)We set up an open system for researchers to upload their own scene maps to obtain customized data sets.
The closed beta test will start after suggestion collection.
2021-07-18 ：
1)Expend our dataset with more CSI data with different SNR levels noise.
2)Publish map files for Scenario 1 indoor office.
Colsed beta test is running.
In the first phase, we plan to provide three researchers (groups) with a full version of dataset generation and 864 core/hours of computing resources. You can use CAD software to make custom map files and save them in '.stl' format. Supported scenarios include, but are not limited to, typical 5G positioning scenarios such as enclosed indoors, city canyons, etc., which should not exceed 1,000 square meters in area.
In addition, you can customize the location, number, and other specific parameters of the base stations and UEs in the map, such as carrier frequency, number of antennas, and bandwidth. If you don't know the specific parameters, you can just submit the map file, and we'll generate your custom dataset based on the default parameters.
Customized datasets with fine-grained CSI for each point and their detailed documentation will be returned after they are generated.
To get your dataset for 5G NR Positioning, please contact us by email. We will start your dataset-generation after confirming your identity and requirements.
Dataset Files
Photo for Scenario 1 indoor office pic_indoor office.png (1,011.48 kB)
Map file for Scenario 1 indoor office mapfile_indoor office.zip (18.52 kB)
Dataset file for Scenario 1 indoor office at an SNR of 50 dB dataset_SNR50_indoor_21-11-16_23-11.mat (537.41 MB)- Dataset file for Scenario 1 indoor office at an SNR of 20 dB dataset_SNR20_indoor_21-11-17_01-50.mat (539.24 MB)
- Dataset file for Scenario 1 indoor office at an SNR of 10 dB dataset_SNR10_indoor_21-11-17_03-07.mat (540.11 MB)
- Photo for Scenario 2 urban canyon pic_outdoor urban canyon.png (128.56 kB)
- feature_outdoor_gNB4-UE1.png (69.71 kB)
- feature_outdoor_gNB4-UE2.png (68.81 kB)
- feature_outdoor_gNB4-UE3.png (69.13 kB)
- Dataset file for Scenario 2 outdoor urban canyon at an SNR of 50 dB dataset_SNR50_outdoor.mat (346.76 MB)
- Dataset file for Scenario 2 outdoor urban canyon at an SNR of 20 dB dataset_SNR20_outdoor.mat (347.77 MB)
- Dataset file for Scenario 2 outdoor urban canyon at an SNR of 10 dB dataset_SNR10_outdoor.mat (348.16 MB)
- An example of data import for Python users. import_5GCSI_dataset_from_mat_file.py.zip (859 bytes)
Open Access dataset files are accessible to all logged in users. Don't have a login? Create a free IEEE account. IEEE Membership is not required.
Comments
where is the paper?
为何数据是实数？？