Preferential attachment alone is not sufficient to generate scale free random networks

TL;DR

This paper demonstrates analytically that preferential attachment alone does not produce scale-free networks, as such networks tend to exhibit exponential degree distributions without growth.

Contribution

The study provides explicit formulas showing that fixed-vertex networks with preferential attachment have exponential degree distributions, challenging the idea that preferential attachment alone suffices for scale-free behavior.

Findings

Degree distributions decay exponentially in fixed-vertex networks with preferential attachment

Growth is necessary for scale-free degree distributions

Preferential attachment alone is insufficient for scale-free networks

Abstract

Many networks exhibit scale free behavior where their degree distribution obeys a power law for large vertex degrees. Models constructed to explain this phenomena have relied on preferential attachment where the networks grow by the addition of both vertices and edges, and the edges attach themselves to a vertex with a probability proportional to its degree. Simulations hint, though not conclusively, that both growth and preferential attachment are necessary for scale free behavior. We derive analytic expressions for degree distributions for networks that grow by the addition of edges to a fixed number of vertices, based on both linear and non-linear preferential attachment, and show that they fall off exponentially as would be expected for purely random networks. From this we conclude that preferential attachment alone might be necessary but is certainly not a sufficient condition for generating scale free networks.