SH-SDS: a new static-dynamic strategy for substation host security detection

Yang Diao, Hui Chen, Wei Liu, Abdur RasoolPeerJ Comput. Sci 2024 · DOI

TL;DR

This paper introduces SH-SDS, a novel static-dynamic strategy for detecting security threats on substation hosts within power grids. By combining configuration checks with simulated attack scenarios, SH-SDS significantly improves detection speed and accuracy, addressing the growing cybersecurity risks in digitized power infrastructure.

What Problem Does This Paper Address?

The increasing digitization of power grids, particularly with the advent of "intelligent substations," has brought efficiency but also heightened vulnerability to cyberattacks. These substations, often located in remote areas, are critical nodes that, if compromised, could lead to widespread power outages. Traditional manual security checks are slow, inefficient, and prone to errors, especially when security requirements change. Existing automated solutions often focus on either static configuration analysis or dynamic intrusion detection, but rarely combine both effectively for the specific needs of substation hosts. This paper aims to bridge this gap by proposing a comprehensive static-dynamic approach to automate and enhance substation host security detection.

Methods

The core of the SH-SDS system is its static-dynamic strategy. This approach combines two types of detection:

  • Static Detection: This involves checking the configuration of network security protection parameters. It's like reviewing the blueprints and ensuring all the safety features are correctly installed and set up. The paper details a process for building a static strategy library by collecting and structuring requirements from cybersecurity standards.
  • Dynamic Detection: This simulates actual attack behaviors to test the system's response. It's akin to performing a fire drill or a simulated intrusion to see how the defenses hold up. A dynamic strategy library is also constructed, focusing on simulating real-world attack scenarios.

The system is built around five modules:

  1. Data Generation Module: Collects and indexes security detection items, requirements, and methods from standard documents.
  2. Strategy Generation Module: Uses RG (Requirement Generation) and SG (Strategy Generation) algorithms to create a tailored static-dynamic strategy for a target host by matching its configuration with the libraries.
  3. Detection Module: Executes the generated strategy to perform the security detection.
  4. Report Module: Summarizes the detection results, including system information, IP addresses, and identified problems.
  5. User Interaction Module: Provides a graphical interface for engineers to initiate scans and view reports.

The paper also describes algorithms for building the static and dynamic strategy libraries and for generating the combined detection strategy.

Datasets

The paper utilizes standard documents related to classified protection for cybersecurity in China, specifically mentioning DL/T 2613-2023 and DL/T 2614-2023. These documents serve as the source for extracting security detection requirements and forming the basis of the static and dynamic strategy libraries. No new benchmark datasets were introduced; instead, existing standards were processed and structured.

Key Insights and Ideas

  • The Synergy of Static and Dynamic Detection: The paper strongly argues that neither static nor dynamic detection alone is sufficient. Combining them provides a more robust and comprehensive security posture, addressing both configuration weaknesses and active threats.
  • Automated Strategy Generation: By developing RG and SG algorithms, the system automates the creation of detection strategies, significantly reducing the manual effort and time required for security checks.
  • Standardization and Library Building: The systematic approach to collecting and structuring cybersecurity requirements from official standards into libraries is crucial for the system's effectiveness and maintainability.
  • Addressing the "Human Factor": The proposed automated system aims to minimize human error inherent in manual inspection, leading to more consistent and reliable security assessments.

Key Results

  • Improved Speed and Accuracy: The SH-SDS system demonstrated significantly faster response times compared to manual inspection and OpenSCAP, while achieving high accuracy (near 100% in simulated tests and over 99% in real-world tests).
  • Efficiency: In real-world tests, SH-SDS outperformed manual detection by up to 98% at Substation C, highlighting its efficiency gains.
  • Resource Efficiency: SH-SDS showed smaller software size and lower CPU utilization compared to OpenSCAP.
  • User-Friendliness: The system provides a graphical interface for initiating scans and viewing results, enhancing usability.
  • Robustness: The static-dynamic strategy proved superior to purely static or dynamic strategies in ablation experiments, showing high accuracy across various test conditions.

Limitations and Future Work

The authors acknowledge several limitations:

  • Maintainability: The software is not easy to maintain, and updating the network security detection standard library requires manual intervention.
  • Interface Imperfection: The user interface is described as imperfect in terms of human-computer interaction and convenience.

Future work suggested includes:

  • Enriching the Standard Library: Collecting more host system network security detection projects to expand the standard library.
  • Automated Requirement Extraction: Developing methods to automatically extract the latest network security detection requirements from standards, cases, and reports.
  • Interface Enhancement: Improving the software's interface for better usability.

Summary updated

BibTeX
@article{diao2024sh,
  author = {Yang Diao and Hui Chen and Wei Liu and Abdur Rasool},
  title = {{SH-SDS: a new static-dynamic strategy for substation host security detection}},
  journal = {PeerJ Comput. Sci.},
  year = {2024},
  doi = {10.7717/peerj-cs.2512},
}
Figure