Geographical threshold graphs with small-world and scale-free properties

Area of Science:

• Network Science
• Statistical Physics
• Complex Systems

Background:

• Real-world networks often exhibit properties like short diameters, high clustering, and power-law degree distributions.
• Existing models, such as Barabási-Albert's preferential attachment, explain these properties in growing networks.
• However, not all networks grow, suggesting the need for models that account for non-growing structures with intrinsic vertex properties.

Purpose of the Study:

• To propose a novel geographical network model that does not rely on network growth.
• To incorporate intrinsic vertex weights and spatial proximity as mechanisms for edge formation.
• To demonstrate how this model can reproduce key network characteristics, including power-law degree distributions.

Main Methods:

• Developed a geographical, non-growing network model where edge formation depends on spatial closeness and/or summed vertex weights.
• Generalized existing models like the unit disk graph, Boolean model, and gravity model within this framework.
• Analyzed the model's ability to produce small-world networks and power-law degree distributions under specific configurations.

Main Results:

• The proposed model successfully generates networks with small-world properties and power-law degree distributions.
• It provides a plausible explanation for observed network features in non-growing systems.
• The model establishes connections between geographical factors, vertex weights, and emergent power-law phenomena in networks.

Conclusions:

• A geographical, non-growing network model with vertex weights offers a more versatile explanation for real-world network structures.
• Spatial proximity and vertex weights are significant drivers for network formation and emergent properties.
• The study highlights the interplay between geography, vertex attributes, and power-law distributions in complex networks.