Finite density scaling laws of condensation phase transition in zero-range processes on scale-free networks*

Guifeng Su, Xiaowen Li, Xiaobing Zhang, Yi ZhangChinese 2020 · DOI

TL;DR

This paper investigates how the density of particles affects condensation phase transitions in a specific type of random process (Zero-Range Process) occurring on complex networks with a scale-free structure. They develop analytical models and use simulations to show that particle density, not just network structure, dictates how particles clump together and how the system settles into a stable state.

What Problem Does This Paper Address?

Condensation phase transitions, where a significant fraction of particles suddenly clumps together, are observed in many physical systems, from traffic flow to Bose-Einstein condensates. Understanding these transitions in non-equilibrium systems, especially those with complex structures like networks, is a major challenge. Zero-Range Processes (ZRPs) are a simplified model for studying such phenomena. Previous work on ZRPs on scale-free networks (networks with a few highly connected nodes, like social networks) focused on the role of network topology. However, the impact of particle density on these transitions, particularly in the steady state and during the relaxation process, remained largely unexplored. This paper aims to fill that gap by analytically and numerically investigating how varying particle density influences condensation in ZRPs on scale-free networks.

Methods

The authors employ a combination of theoretical and computational approaches:

  • Grand Canonical Ensemble (GCE) Method: This statistical mechanics framework is used to analytically predict the behavior of the system. It allows for variable particle density by introducing a "fugacity" parameter.
  • Scaling Ansatz and Hierarchical Dynamics: For analyzing the relaxation dynamics (how the system reaches its steady state), they use a scaling ansatz. This assumes that the relaxation process occurs in stages, with smaller clusters merging into larger ones, and that this process can be described by power-law relationships. They also propose a "hierarchical evolution" where nodes with lower degrees equilibrate first, followed by larger ones.
  • Monte Carlo (MC) Simulations: To validate their analytical predictions, they perform extensive MC simulations on scale-free networks generated using the Barabási-Albert model. These simulations track the particle distribution and dynamics over time.

Datasets

Not discussed in the paper. The study uses simulated scale-free networks generated via the Barabási-Albert model.

Key Insights and Ideas

  • Density as a Crucial Factor: The paper strongly emphasizes that particle density is not just a passive parameter but actively influences the condensation transition. It dictates the "crossover degree" (the degree of nodes where the behavior shifts from fluid-like to condensed) and the average occupation of nodes.
  • Novel Scaling Laws: They derive new analytical scaling laws that describe how the crossover degree and average occupation number change with particle density in both the steady state and during relaxation.
  • Hierarchical Relaxation Dynamics: The study supports the idea that relaxation in these complex systems is hierarchical, with smaller, lower-degree nodes equilibrating before larger, higher-degree nodes. This process is also shown to be density-dependent.
  • Validation through Simulation: The Monte Carlo simulations provide strong empirical support for the analytically derived scaling laws, demonstrating the accuracy of their theoretical framework.

Key Results

  • Steady State: The crossover degree ($k_c$) and average occupation number ($m_k$) are found to be density-dependent, exhibiting specific scaling relations with particle density ($\rho$) and the interaction parameter ($\delta$). For instance, $k_c$ scales as $\rho^{-\delta/\delta_c}$ for $\delta < \delta_c$.
  • Relaxation Dynamics: The average occupation number ($m_k$) during transient times scales with both node degree and density, as $m_k \sim k \rho^{\delta_c/(\delta_c-\delta)}$. The ratio of the Inverse Participation Ratio (IPR) to its steady-state limit ($I_t/I_\infty$) scales with time and density as $t/\rho^\nu$, where $\nu = 1-\delta$ is the dynamic exponent.
  • Simulation Confirmation: Figures in the paper show excellent data collapse for simulation results, confirming the predicted scaling behaviors for both steady-state and relaxation dynamics across various densities and network sizes.

Limitations and Future Work

  • Specific Network Model: The study focuses on scale-free networks generated by the Barabási-Albert model. While representative, other network generation models might yield different quantitative results.
  • Interaction Parameter Range: The analysis primarily focuses on the case $0 < \delta < 1$, where attractive interactions are present. The behavior for repulsive interactions ($\delta > 1$) is briefly mentioned but not explored in detail.
  • Future Work: The authors suggest further exploration of the relaxation dynamics, particularly the role of the dynamic exponent and the influence of different network structures. Investigating the impact of finite system size on the scaling laws is also a natural extension.

Summary updated

Abstract

The dynamics of zero-range processes on complex networks is expected to be influenced by the topological structure of underlying networks. A real space complete condensation phase transition in the stationary state may occur. We study the finite density effects of the condensation transition in both the stationary and dynamical zero-range processes on scale-free networks. By means of grand canonical ensemble method, we predict analytically the scaling laws of the average occupation number with respect to the finite density for the steady state. We further explore the relaxation dynamics of the condensation phase transition. By applying the hierarchical evolution and scaling ansatz, a scaling law for the relaxation dynamics is predicted. Monte Carlo simulations are performed and the predicted density scaling laws are nicely validated.
BibTeX
@inproceedings{su2020finite,
  author = {Guifeng Su and Xiaowen Li and Xiaobing Zhang and Yi Zhang},
  title = {{Finite density scaling laws of condensation phase transition in zero-range processes on scale-free networks*}},
  booktitle = {Chinese},
  year = {2020},
  doi = {10.1088/1674-1056/ab8a41},
}

Loading related content...

Figure