search

aektasharma / Veremi-dataset-classification

Classification of all five types of position falsification attack present in VeReMI dataset.
GNU General Public License v3.0
6 stars 6 forks source link

A small problem #1

Open 17860511195 opened 1 year ago

17860511195 commented 1 year ago

Hello, I am a student, currently very interested in the Internet of Vehicles detection, I was very inspired to read the code you shared. Is the csv table in the dataset folder in your shared library all the data of the VeReMi dataset? I can only download the source code to the virtual environment elsewhere and don't know how to export the data. If yes, it's fantastic! I hope you can answer my question, it is important to me, thank you very much again.

aektasharma commented 1 year ago

Hello, The dataset folder is not the original Veremi Dataset. You can download the data from here: https://veremi-dataset.github.io/ Veremi Dataset contained a lot of fields (noise) which were not useful for my implementation so I had structured, cleaned and reproduced the dataset based on the fields and data I required. There is another csv called [modifiedat16.csv](https://github.com/aektasharma/Veremi-dataset-classification/blob/main/Dataset/modifiedat16.csv). This was generated by me using the original attack-16 data, for more information on this work, you can refer to our paper: A. Sharma and A. Jaekel, "Machine Learning Based Misbehaviour Detection in VANET Using Consecutive BSM Approach," in IEEE Open Journal of Vehicular Technology, vol. 3, pp. 1-14, 2022, doi: 10.1109/OJVT.2021.3138354 Thank you!

17860511195 commented 1 year ago

Thank you very much for your reply, I will study this paper carefully.

17860511195 commented 1 year ago

Hello, teacher, the link to the paper you sent me last time, I have read the paper and code, very rewarding. There are a few questions I would like to ask. First, is the dataset you created extracted from each GroundTruthJSONlog.json file in the VeReMi dataset? The second is that the core of this paper is to describe that two continuous BSMs as a sample are better than ordinary direct use of BSM, so how to deal with discontinuous situations, such as this data set is just odd, and it must be the last one BSM has no BSM paired with it. The third is how to determine the label? Is it determined by the fact that the two BSMs belong to the same category? What if two consecutive BSMs belong to different categories? Very much looking forward to your reply. Student Li.