A Federated Capability-based Access Control Mechanism for Internet of Things (IoTs)

Ronghua Xu, Yu Chen, Erik Blasch, Genshe Chen

TL;DR

This paper introduces FedCAC, a novel federated capability-based access control framework designed for the massive, heterogeneous, and dynamic nature of IoT environments. By distributing authorization decisions to local "coordinators" and using lightweight capability tokens, FedCAC aims to provide a scalable, fine-grained, and efficient security solution for IoT devices.

What Problem Does This Paper Address?

The Internet of Things (IoT) is rapidly expanding, connecting a vast number of diverse devices that collaboratively offer intelligent services. While this brings immense benefits, it also introduces significant security and privacy concerns. A critical challenge is access authorization: controlling who can access what resources and services within these complex networks. Traditional access control models like Access Control Lists (ACLs), Role-Based Access Control (RBAC), and Attribute-Based Access Control (ABAC) struggle to cope with the sheer scale, heterogeneity, and dynamic nature of IoT systems. They often lead to management overload, inflexibility, and are not lightweight enough for resource-constrained IoT devices. This paper addresses the need for a more scalable, manageable, and efficient access control mechanism tailored for the unique demands of IoT.

Methods

The core of the proposed solution is the Federated Capability-based Access Control (FedCAC) framework. FedCAC operates on a three-tiered architecture:

  • Cloud Computing Layer: Handles global management, including a central Policy Decision Making center (PDC) for identity and policy rule management.
  • Fog Computing Layer: Features "coordinators" (e.g., laptops, smartphones) that delegate domain-specific authorization tasks from the PDC.
  • Edge Computing Layer: Consists of service providers (IoT devices) that process access control requests locally.

Key mechanisms include:

  • Identity-based Capability Tokens: Entities are assigned unique Virtual IDs (VIDs). Capability tokens, represented in JSON format for lightweightness, encapsulate access rights and are signed by the issuer.
  • Federated Delegation: The PDC delegates authorization responsibilities to domain coordinators, distributing the workload and enabling local policy enforcement.
  • Capability Propagation and Validation: When a user requests access, a capability token is issued. This token is then presented to the service provider, which validates its authenticity, access rights, and any contextual conditions before granting access.
  • Decentralized Authorization: Service providers at the edge perform local access right validation, integrating situational awareness and contextual conditions.

Datasets

Not discussed in the paper. The paper focuses on a conceptual prototype and experimental validation using a simulated IoT network.

Key Insights and Ideas

  • Shifting the Burden: The most significant insight is the shift of authorization decision-making from a centralized cloud to distributed fog and edge nodes. This "federated" approach is crucial for scalability in IoT.
  • Lightweight Tokens are Key: Using JSON for capability tokens is a deliberate choice to minimize overhead on resource-constrained IoT devices, a common bottleneck in existing solutions.
  • Delegation as a Scalability Enabler: The hierarchical delegation mechanism, from PDC to coordinators, is central to managing the complexity and scale of IoT access control.
  • Context Matters at the Edge: Empowering edge devices to enforce access control based on local context (situational awareness) allows for more dynamic and fine-grained security policies.

Key Results

The experimental results demonstrate the feasibility and effectiveness of the FedCAC scheme.

  • Network Latency: FedCAC introduces an average of 11 ms of extra latency compared to a benchmark without capability-based access control, which is considered acceptable (a ~4% increase over a 300 ms baseline).
  • Execution Time: The overall execution time for retrieving data with FedCAC is approximately 42.57 ms. The authorization process itself takes about 7.823 ms, with issuer signature verification being the most computationally intensive step (around 75% of the authorization time).
  • Feasibility: The prototype was successfully implemented and tested on resource-constrained devices like Raspberry Pi, showcasing its suitability for such environments.

Limitations and Future Work

The authors acknowledge several areas for future research:

  • Intelligence on Edge Networks: While delegation brings some intelligence to the edge, the core policy decision-making and identity management remain centralized. Future work aims to further decentralize this intelligence, allowing devices to have more autonomy.
  • Decentralized Authentication and Authorization: Investigating blockchain technology for decentralized authentication and authorization is proposed to address trustless network environments and enhance anonymity.
  • Data Heterogeneity: Integrating IoT data with other sources like multimedia, text, and social data is a future direction to enrich access control policies.
  • Inter-domain Scenarios: The current experiments were conducted within a single network domain. Future work will explore the impact of FedCAC in inter-domain scenarios, where network latency (RTT) will be a more significant factor.

Summary updated

Abstract

The prevalence of Internet of Things (IoTs) allows heterogeneous embedded smart devices to collaboratively provide intelligent services with or without human intervention. While leveraging the large-scale IoT-based applications like Smart Gird and Smart Cities, IoT also incurs more concerns on privacy and security. Among the top security challenges that IoTs face is that access authorization is critical in resource and information protection over IoTs. Traditional access control approaches, like Access Control Lists (ACL), Role-based Access Control (RBAC) and Attribute-based Access Control (ABAC), are not able to provide a scalable, manageable and efficient mechanisms to meet requirement of IoT systems. The extraordinary large number of nodes, heterogeneity as well as dynamicity, necessitate more fine-grained, lightweight mechanisms for IoT devices. In this paper, a federated capability-based access control (FedCAC) framework is proposed to enable an effective access control processes to devices, services and information in large scale IoT systems. An identity-based capability token management strategy is presented, which involves registering, propagation and revocation of the access authorization. Through delegating centralized authorization decision-making policy to local domain delegator, the access authorization process is locally conducted on the service provider that integrates situational awareness (SAW) and customized contextual conditions. Implemented and tested on both resources-constrained devices, like smart sensors and Raspberry PI, and non-resource-constrained devices, like laptops and smart phones, our experimental results demonstrate the feasibility of the proposed FedCAC approach to offer a scalable, lightweight and fine-grained access control solution to IoT systems connected to a system network.
BibTeX
@misc{xu2018federated,
  author = {Ronghua Xu and Yu Chen and Erik Blasch and Genshe Chen},
  title = {{A Federated Capability-based Access Control Mechanism for Internet of Things (IoTs)}},
  year = {2018},
  eprint = {1805.00825},
  archivePrefix = {arXiv},
  url = {https://arxiv.org/abs/1805.00825},
}
Figure