Performance Analysis of Dynamic ABAC Systems using a Queuing Theoretic Framework

Area of Science:

• Computer Science
• Information Security
• Systems Engineering

Background:

• Attribute-Based Access Control (ABAC) systems rely on policies for access decisions.
• Dynamic ABAC systems require managing authorization changes via auxiliary lists, impacting efficiency.
• Periodic policy rebuilding is necessary but computationally expensive and causes system unavailability.

Purpose of the Study:

• To address the challenge of balancing access request resolution, dynamic authorization updates, and ABAC policy rebuilding.
• To investigate performance trade-offs in dynamic ABAC systems.

Main Methods:

• Modeled access mediation as an M/G/1 queue with vacation/limited service.
• Analyzed the impact of queue discipline on performance parameters.
• Conducted extensive experiments to evaluate system behavior.

Main Results:

• The queuing model effectively represents the trade-offs in dynamic ABAC systems.
• Queue discipline significantly influences access resolution time and system availability.
• Experimental results offer guidance for efficient dynamic ABAC implementation.

Conclusions:

• A queuing theoretic approach provides a framework for optimizing dynamic ABAC systems.
• Careful consideration of queue discipline is crucial for performance.
• The findings facilitate more efficient and available dynamic ABAC systems.