Datasets
Standard Dataset
Synthetic dataset for mobile wireless networks with SUMO -- individual traces
- Citation Author(s):
- Submitted by:
- Afonso Oliveira
- Last updated:
- Mon, 09/27/2021 - 17:39
- DOI:
- 10.21227/4zdn-qe46
- Data Format:
- License:
914 Views
- Categories:
- Keywords:
0 ratings - Please login to submit your rating.
Abstract
This dataset contains individual traces of a published synthetic dataset generator for wireless mobile networks using SUMO. The generated dataset is done with the map of the city of Berlin as base. The dataset has 21 cells of 7 base stations. Separate files were generated for traffic and radio signal quality, but can be joined by using the UE identifier.
Instructions:
Files are in a compressed parquet format. The easiest way to use them is to use Pandas library for Python.
import pandas as pd
pd.read_parquet(<file>)
Comments
hello, is there any information about the time interval of the measurement?
Thanks.
Hello Tingting Liu,
The interval of SNR traces is 1ms. Just to remind, this is a synthetically generated dataset. The traffic traces are generated each time a packet is triggered. If it fits your needs, the aggregated version of this dataset that aggregates SNR and traffic traces in time windows of 10s (uses considerably less space) is available in Zenodo: https://zenodo.org/record/5513705.
The generator is available in GitHub: https://github.com/afonsomota/5gdatasetgenerator. In the repository, you have a PDF with some descriptions of the generator procedures.
The generator and traces are part of an accepted article to a conference, but the presentation has not been made yet at the time of writing this answer.
Best regards.
hello Afonso Oliveira,
this reply is really helpful. Thank you so much.
I have another question, what's the relationship between the 7 base station and the 21 cells,like if a user is belong to cell 18,which BS is it belong to? in this dataset, is the BS number 7?
And the dataset in Zenodo, what's the difference of the the three files "aggregated-t_10-att_0.parquet","aggregated-t_10-att_10.parquet"and "aggregated-t_10-att_30.parquet".
Look forward to your reply.
Best Regards.
Thank you.
Hello Cecilia Zhao,
I have just noticed that the article is now online: https://dl.acm.org/doi/abs/10.1145/3479241.3486704.
Figures 2 and 3 answer your question related to the base stations and cells.
The aggregated traces difference is in the additional attenuation given to the UEs. You can find more information in Section 4.3.1 of the paper.
Best Regards.
Hello Cecilia Zhao,
The different datasets have different attenuation factors applied to the nodes besides the ones generated. If you reach me by e-mail (afonso.oliveira@tecnico.ulisboa.pt), I can send you the article presentation that has more information. It includes illustrations of which cell belongs to each base station and where they are located in relation to each other.
Best Regards.
Does your dataset contain base station centric data? I couldn't find Readme file explaining the features collected