Scaling phenomena in the Internet: critically examining criticality

Area of Science:

• Network Science
• Complex Systems
• Internet Measurement

Background:

• Internet traffic exhibits self-similar scaling in burst patterns.
• Internet topology displays scale-free structure in some contexts.
• Emergent phenomena in complex systems are often explained using concepts like fractals and criticality.

Purpose of the Study:

• To evaluate criticality-based models explaining Internet scaling phenomena.
• To distinguish between evocative models and truly explanatory models.
• To establish a framework for developing verifiable theories of large networks.

Main Methods:

• Developed a validation framework to test model causality.
• Applied the framework to assess criticality-based explanations of Internet scaling.
• Examined both traffic flow and network topology scaling behaviors.

Main Results:

• Criticality-based models can reproduce observed emergent scaling phenomena.
• These models fail to capture the true underlying causal mechanisms.
• The proposed validation framework offers a basis for network theory development.

Conclusions:

• Criticality-based models are insufficient as explanations for Internet scaling.
• A more rigorous approach is needed to understand the 'why' and 'how' of Internet scaling.
• The validation framework provides a path toward a verifiable theory of large networks.