SOCO2023:

Abdelhakim Baouya, Brahim Hamid, Levent Gürgen, and Saddek Bensalem Formal Modeling and Analysis of Tampering Attacks and Energy Consumption Effects on Edge Servers using Concurrent Stochastic Games. In Soft Computing, Springer , 2023 (Submitted 2023).

Developed Artefacts

ID	PRISM-games Model Description	Model	rPATL properties
M1	Example of simple CSG game where two players are producer and consumer	[Download]	[Download]
M2	Model of Sensor-Edge-Cloud Architecture under attack implementing Synchronous Mode	[Download]	[Download]
M3	Model of Sensor-Edge-Cloud Architecture under attack implementing Asynchronous Mode	[Download]	[Download]
M4	[Mitigation] Model of Sensor-Edge-Cloud Architecture under attack with defense Startegy	[Download]	[Download]
M4	[Scalability] Model of Sensor-Edge-Cloud Architecture under attack implementing Synchronous Mode with Three Attacks	[Download]	[Download]
M5	[Scalability] Model of Sensor-Edge-Cloud Architecture under attack implementing Asynchronous Mode with Three Attacks	[Download]	[Download]
ID	Python Code Description	Code	Dataset Link
C1	Python code for estimation of attack rate (details below) Please disregard the error message on the Dataset link. To obtain the file, download it by clicking on the "Telecharger/Download" button located in the top right corner	[Download]	[Link]

Estimation of Attacks Rate

This code aims to estimate the mean time between attacks and the attack rate using different methods is licensed under CeCILL-B.. The provided command should be executed to obtain the results.

Execution Command

To execute the command, run the following command in your terminal:

python3 main.py ARP data.csv 6 84

This command takes the following parameters:

ARP: Attack type
data.csv: CSV file containing the attack data
6: Column index of the time column in the CSV file
84: Column index of the attack column in the CSV file

Results

After executing the command, the program will output the following results:

Attack: ARP
CSV File: data.csv
Time Column: 6
Attack Column: 84

The estimated mean time between attacks is 7.414857530529172 seconds.

The estimated rate parameter using the classical method is 0.13486435793037194.

The estimated rate parameter using Maximum Likelihood Estimation (MLE) with the solver L-BFGS-B is 0.13486435351259937.

The estimated rate parameter using MLE with the Nelder-Mead solver is 0.13486328124999925.

To reuse the contact please contact me.