Survey on Enterprise Internet-of-Things systems (E-IoT): A security perspective

Luis Puche Rondon, Leonardo Babun, Ahmet Aris, Kemal Akkaya, A. Selcuk Uluagac

TL;DR

This paper provides the first comprehensive survey of security threats and defenses specifically for Enterprise Internet-of-Things (E-IoT) systems. By dissecting E-IoT into four distinct layers (Devices, Communications, Monitoring/Applications, and Business), the authors systematically analyze vulnerabilities and propose mitigation strategies, highlighting the critical need for greater awareness and research in this often-overlooked area.

What Problem Does This Paper Address?

While the security of consumer-grade IoT devices has garnered significant attention, Enterprise IoT (E-IoT) systems, which power more sophisticated and often high-profile deployments (e.g., smart buildings, luxury homes, secure conference rooms), have been largely overlooked. These E-IoT systems, characterized by their proprietary nature, customizability, and complexity, are often mistakenly assumed to be secure due to their cost and specialized installation. This lack of security research creates a significant vulnerability, as E-IoT systems are increasingly targeted. This paper aims to bridge this gap by providing a structured overview of E-IoT components, their associated threats, and existing or potential defense mechanisms.

Methods

The authors employ a layered approach to systematically analyze E-IoT security. They divide E-IoT systems into four distinct layers:

  1. E-IoT Devices Layer: Focuses on physical components, sensors, and their direct interactions.
  2. Communications Layer: Examines the protocols and interfaces used for device communication.
  3. Monitoring and Applications Layer: Covers software-based components, drivers, and system configuration.
  4. Business Layer: Addresses cloud services, vendor services, and remote access functionalities.

For each layer, the paper identifies components, outlines threat models (assuming an attacker named "Mallory"), details specific attacks and vulnerabilities, and discusses relevant defense and mitigation strategies. This structured approach allows for a comprehensive and organized examination of the E-IoT security landscape.

Datasets

Not discussed in the paper.

Key Insights and Ideas

  • E-IoT is Not Inherently Secure: Despite their complexity and cost, E-IoT systems are not immune to attacks. The proprietary nature and "security through obscurity" approach often employed by vendors can mask significant vulnerabilities.
  • Layered Security is Crucial: A holistic security approach is necessary, as attackers can exploit vulnerabilities across multiple layers (cross-layer attacks).
  • Proprietary Nature is a Double-Edged Sword: While proprietary systems offer customization, their closed-source nature hinders security research and makes it difficult to identify and address vulnerabilities.
  • Integrators are a Critical Link (and Potential Weakness): The reliance on specialized integrators for E-IoT deployment and maintenance introduces a significant trust factor. Malicious or careless integrators can inadvertently compromise entire systems.
  • Legacy Systems Pose a Unique Challenge: Many E-IoT systems are built on older infrastructure that may no longer receive security updates, creating persistent vulnerabilities.
  • Supply Chain Attacks are a Growing Concern: The manufacturing and distribution stages of E-IoT components offer ample opportunities for attackers to introduce malicious hardware or software.

Key Results

The paper provides a detailed taxonomy of attacks and defenses across the four identified layers of E-IoT systems. While not presenting quantitative results from new experiments, it synthesizes existing research and identifies critical security gaps. Key findings include:

  • Device Layer: Attacks like supply chain compromises, physical access exploits, and side-channel attacks are prevalent. Mitigations involve secure supply chain practices, physical security, and robust authentication.
  • Communications Layer: Vulnerabilities exist in proprietary protocols, WiFi, ZigBee/Z-Wave, and Bluetooth. Defenses include using secure protocols (WPA3, updated ZigBee/Z-Wave implementations), encryption, and network segmentation.
  • Monitoring and Applications Layer: Driver-based attacks, software vulnerabilities, and insecure configurations are major threats. Recommendations include using verified drivers, frequent software updates, and secure configuration practices.
  • Business Layer: Cloud service vulnerabilities, data breaches, and account hijacking are significant concerns. Mitigation strategies involve strong encryption, access control, and secure cloud service management.

Limitations and Future Work

The authors acknowledge several limitations and identify areas for future research:

  • Lack of Standardization: The proprietary nature of E-IoT systems makes it difficult to develop universal security solutions.
  • Limited Vendor Cooperation: The closed-source nature of many E-IoT components hinders independent security research.
  • Need for Auditing Tools: There's a need for better auditing mechanisms to ensure E-IoT systems are properly configured and secure.
  • Protection for Legacy Systems: Developing solutions to protect E-IoT systems that cannot be updated is a critical area for future work.
  • Cross-Layer Attack Defense: More research is needed on defending against sophisticated cross-layer attacks.
  • Honeypots for E-IoT: Further exploration of honeypot strategies tailored for E-IoT environments could provide valuable insights into attacker behavior.
  • Privacy-Preserving Techniques: Research into advanced encryption methods like Homomorphic Encryption for cloud-based E-IoT data storage is promising but requires further development.

Summary updated

BibTeX
@article{rondon2022survey,
  author = {Luis Puche Rondon and Leonardo Babun and Ahmet Aris and Kemal Akkaya and A. Selcuk Uluagac},
  title = {{Survey on Enterprise Internet-of-Things systems (E-IoT): A security perspective}},
  journal = {Ad Hoc Networks},
  year = {2022},
  doi = {10.1016/j.adhoc.2021.102728},
}
Figure