Random Networks with given Rich-club Coefficient

TL;DR

This paper introduces a new method for generating surrogate networks that preserve the rich-club coefficient, enabling more accurate modeling of real networks' density and mixing patterns.

Contribution

It presents a novel network model based on conserving the rich-club coefficient, offering an alternative to degree-based models.

Findings

The method accurately reproduces degree distribution and mixing patterns.

Generated networks match real network properties.

The approach is simple to implement and effective.

Abstract

In complex networks it is common to model a network or generate a surrogate network based on the conservation of the network's degree distribution. We provide an alternative network model based on the conservation of connection density within a set of nodes. This density is measure by the rich-club coefficient. We present a method to generate surrogates networks with a given rich-club coefficient. We show that by choosing a suitable local linking term, the generated random networks can reproduce the degree distribution and the mixing pattern of real networks. The method is easy to implement and produces good models of real networks.