smartest: aSsessMent frAmework foR deTEctors of integrity attacks to Smart meTers

This repository includes an experimental framework for assessing the quality and performance of detectors of integrity attacks to smart-meters.

Structure of the repository

• src: this folder includes the Python source code (organized in sub-folders):
  • preprocessing: datasets pre-processing module and synthetic dataset generator
  • detectors: software detectors, detector infrastructure
  • experiments: the experiments launcher module
  • analytics: the dashboard module
• script_results: outputs from the experiments.
• requirements.txt: the Python3 packages required to run the scripts
• README: this file
• .gitignore: specifies intentionally untracked files to ignore

Usage

The dependencies of the tool are listed in requirements.txt. You can run the following command for installing them:

$ pip install -r requirements.txt

You must run all the python3 commands from the root directory.

Reference

The original input datasets are the electricity and gas dataset from the Irish Social Science Data Archive - Commision for Energy Regulation (ISSDA-CER). Both are available, for research purposes, upon request, and the information provided is anonymous.

• External link: https://www.ucd.ie/issda/data/commissionforenergyregulationcer/

The experimental framework and experiments results are detailed in:

• S. Bernardi, R. Javierre, J. Merseguer, J. Requeno, Detectors of Smart Grid Integrity Attacks: An Experimental Assessment, 17th European Dependable Computing Conference, 13-16 September 2021, Munich, Germany. See preprint.