Evolution of reference networks with aging

Abstract

We study the growth of a reference network with aging of sites defined in the following way. Each new site of the network is connected to some old site with probability proportional (i) to the connectivity of the old site as in the Barabási-Albert's model and (ii) to $τ^{-α}$, where $τ$ is the age of the old site. We consider $α$ of any sign although reasonable values are $0 \leq α\leq \infty$. We find both from simulation and analytically that the network shows scaling behavior only in the region $α< 1$. When $α$ increases from $-\infty$ to 0, the exponent $γ$ of the distribution of connectivities ($P(k) \propto k^{-γ}$ for large $k$) grows from 2 to the value for the network without aging, i.e. to 3 for the Barabási-Albert's model. The following increase of $α$ to 1 makes $γ$ to grow to $\infty$. For $α>1$ the distribution $P(k)$ is exponentional, and the network has a chain structure.