The effect of randomness for dependency map on the robustness of interdependent lattices

Jing Yuan, Lixiang Li, Haipeng Peng, Jürgen Kurths, Xiaojing Hua, Yixian Yang

Abstract

For interdependent networks with identity dependency map, percolation is exactly the same with that on a single network and follows a second-order phase transition, while for random dependency, percolation follows a first-order phase transition. In real networks, the dependency relations between networks are neither identical nor completely random. Thus in this paper, we study the influence of randomness for dependency maps on the robustness of interdependent lattice networks. We introduce approximate entropy ($ApEn$) as the measure of randomness of the dependency maps. We find that there is critical $ApEn_c$ below which the percolation is continuous, but for larger $ApEn$, it is a first-order transition. With the increment of $ApEn$, the $p_c$ increases until $ApEn$ reaching $ApEn_c'$ and then remains almost constant. The time scale of the system shows rich properties as $ApEn$ increases. Our results uncover that randomness is one of the important factors that lead to cascading failures of spatially interdependent networks.

Introduction

Modern infrastructure—like power grids, water systems, and communication networks—is increasingly interdependent. A failure in one system can trigger a cascading collapse in another, leading to total system failure. Historically, research has focused on two extremes: "identity maps," where nodes depend on their direct counterparts in another layer, and "totally random maps." However, real-world dependencies are often semi-ordered. This study bridges the gap by investigating how the degree of randomness in dependency links affects the collective stability of interdependent lattice networks.

Methods

The authors model a system of two interdependent $L \times L$ square lattices. To transition from order to randomness, they employ a rewiring strategy inspired by the Watts-Strogatz model. Starting with an identity map, each dependency link is rewired with a probability $q$ (ranging from 0 to 1).

Dependency maps between two lattice layers with increasing randomness from (a) to (d).

To quantify this randomness, the paper introduces Approximate Entropy (ApEn) as a metric for the complexity of the dependency sequence. Robustness is evaluated by removing a fraction of nodes ($1-p$) and tracking the "mutually connected giant component" ($P_\infty$). The time scale of collapse is measured by the Number of Iterations (NOI) required for the cascade to stabilize.

Results

The study identifies a critical threshold of randomness ($q_c \approx 0.13$ for lattices). Below this threshold, the system undergoes a second-order (continuous) phase transition, where the network degrades gracefully. Once randomness exceeds $q_c$, the system becomes brittle, exhibiting a first-order (discontinuous) phase transition characterized by sudden, total collapse.

Size of the giant component $P_\infty$ versus node survival probability $p$ for different rewiring probabilities $q$, showing the transition from continuous to discontinuous percolation.

Key findings include:

  • Percolation Shift: As $q$ increases, the percolation threshold $p_c$ increases, meaning more random systems are more vulnerable to initial failures.
  • Randomness vs. Distance: Randomness in the dependency map is a more critical driver of vulnerability than the physical length of dependency links. Even short-range dependencies can be catastrophic if they are sufficiently random.
  • Topology Sensitivity: Spatially embedded lattices are found to be more sensitive to dependency randomness than Scale-Free (SF), Erdős-Rényi (ER), or Watts-Strogatz (WS) networks.

Discussion and Conclusion

The results demonstrate that randomness in dependency maps is a primary factor leading to catastrophic cascading failures in spatially interdependent networks. The "Number of Iterations" (NOI) peaks at the critical threshold, suggesting that systems near collapse take longer to stabilize—a potential early-warning signal. Future work could explore how these findings apply to dynamical processes like synchronization or disease spread, and how intentionally ordering dependency maps might enhance infrastructure resilience.

Summary updated

Abstract

For interdependent networks with identity dependency map, percolation is exactly the same with that on a single network and follows a second-order phase transition, while for random dependency, percolation follows a first-order phase transition. In real networks, the dependency relations between networks are neither identical nor completely random. Thus in this paper, we study the influence of randomness for dependency maps on the robustness of interdependent lattice networks. We introduce approximate entropy($ApEn$) as the measure of randomness of the dependency maps. We find that there is critical $ApEn_c$ below which the percolation is continuous, but for larger $ApEn$, it is a first-order transition. With the increment of $ApEn$, the $p_c$ increases until $ApEn$ reaching ${ApEn}_c'$ and then remains almost constant. The time scale of the system shows rich properties as $ApEn$ increases. Our results uncover that randomness is one of the important factors that lead to cascading failures of spatially interdependent networks.
BibTeX
@inproceedings{yuan2016effect,
  author = {Jing Yuan and Lixiang Li and Haipeng Peng and Jürgen Kurths and Xiaojing Hua and Yixian Yang},
  title = {{The effect of randomness for dependency map on the robustness of interdependent lattices}},
  booktitle = {Chaos},
  year = {2016},
  doi = {10.1063/1.4939984},
}
Figure