Self-* Capabilities of Cloud-Edge Nodes: A Research Review

Area of Science:

• Computer Science
• Distributed Systems
• Network Computing

Background:

• Edge and fog computing architectures extend cloud-native features to network peripheries, reducing latency and resource usage.
• Autonomous management of these distributed systems requires self-* capabilities in computing nodes to minimize human intervention.
• A systematic classification and analysis of these self-* capabilities are currently lacking.

Purpose of the Study:

• To conduct a literature review on self-* capabilities essential for autonomous edge and fog computing systems.
• To propose a unifying taxonomy for these capabilities in a heterogeneous field.
• To analyze implementation challenges and the absence of a reference architecture.

Main Methods:

• Systematic literature review of self-* capabilities in edge and fog computing.
• Analysis of existing research to identify common and distinct self-* traits.
• Synthesis of findings to propose a potential taxonomy.

Main Results:

• Identified key self-* capabilities required for autonomous operation in edge and fog environments.
• Highlighted the heterogeneity in how these capabilities are addressed and their case-specific dependencies.
• Confirmed the absence of a clear reference architecture for equipping nodes with necessary traits.

Conclusions:

• A unifying taxonomy for self-* capabilities in autonomous edge and fog computing is proposed.
• The heterogeneous implementation and case-specific nature of these capabilities present challenges.
• There is a need for a guiding reference architecture to inform the development of autonomous edge and fog systems.