Abstract
We investigate the sensitivity of epidemic behavior to a bounded susceptibility constraint -- susceptible nodes are infected by their neighbors via the regular SI/SIS dynamics, but subject to a cap on the infection rate. Such a constraint is motivated by modern social networks, wherein messages are broadcast to all neighbors, but attention spans are limited. Bounded susceptibility also arises in distributed computing applications with download bandwidth constraints, and in human epidemics under quarantine policies. Network epidemics have been extensively studied in literature; prior work characterizes the graph structures required to ensure fast spreading under the SI dynamics, and long lifetime under the SIS dynamics. In particular, these conditions turn out to be meaningful for two classes of networks of practical relevance -- dense, uniform (i.e., clique-like) graphs, and sparse, structured (i.e., star-like) graphs. We show that bounded susceptibility has a surprising impact on epidemic behavior in these graph families. For the SI dynamics, bounded susceptibility has no effect on star-like networks, but dramatically alters the spreading time in clique-like networks. In contrast, for the SIS dynamics, clique-like networks are unaffected, but star-like networks exhibit a sharp change in extinction times under bounded susceptibility. Our findings are useful for the design of disease-resistant networks and infrastructure networks. More generally, they show that results for existing epidemic models are sensitive to modeling assumptions in non-intuitive ways, and suggest caution in directly using these as guidelines for real systems.
Introduction
Classical epidemic models, such as the Susceptible-Infected (SI) and Susceptible-Infected-Susceptible (SIS) models, typically assume that a node's risk of infection scales linearly with the number of its infected neighbors. While mathematically convenient, this assumption often fails to capture real-world constraints like limited attention spans in social media or bandwidth limits in distributed computing. This paper introduces the concept of bounded susceptibility, where the infection rate is capped by a maximum value regardless of neighborhood size. This realistic constraint fundamentally alters network vulnerability, often turning influential "hubs" into bottlenecks.
Methodology and Network Models
The study formalizes bounded susceptibility through an infection-profile function $\Phi(\cdot)$ that is concave and bounded. Two primary dynamics are analyzed:
- $SI_{bs}$: Nodes remain infected indefinitely; the metric of interest is the spreading time required to infect the entire network.
- $SIS_{bs}$: Nodes can recover; the metric of interest is the extinction time (duration of infection).
The analysis focuses on two distinct network topologies:
- Clique-like: Dense networks with high connectivity (e.g., Erdős-Rényi graphs).
- Star-like: Sparse networks characterized by a central hub and numerous leaves (e.g., hub-and-spoke infrastructure).
Theoretical bounds are derived using Continuous-Time Markov Chains (CTMC) and isoperimetric constants, validated through simulations on complete graphs, star graphs, and power-law networks.
Results
SI Dynamics: Spreading Time in Dense Networks
In standard SI models, spreading time in a clique vanishes as the network size $n$ grows ($O(\frac{\log n}{n})$). However, under bounded susceptibility ($SI_{bs}$), the spreading time increases logarithmically with $n$ ($\Omega(\log n)$). This dramatic shift occurs because the cap on infection rates prevents the exponential acceleration typically seen in dense graphs.
SIS Dynamics: Extinction Time in Star Networks
In star-like networks, the impact is reversed. In standard SIS models, high-degree hubs sustain the infection, leading to exponential extinction times. Under bounded susceptibility ($SIS_{bs}$), these hubs become bottlenecks. Because their infection rate is capped, they cannot effectively re-infect their neighbors before they recover, causing the extinction time to drop to polynomial scaling.
Discussion and Implications
The findings reveal a "topological sensitivity" where the impact of bounded susceptibility depends heavily on the underlying graph structure and the specific epidemic model.
- The Hub Paradox: Hubs, traditionally seen as engines of persistence, become bottlenecks under bounded susceptibility in SIS models.
- Scale-Free Networks: Despite the hub bottleneck, power-law networks still exhibit long-lasting SIS epidemics because they contain dense, clique-like subgraphs that sustain the infection locally.
- Policy Design: The results suggest that "partial quarantine" (capping susceptibility) is highly effective for star-like infrastructure (e.g., air travel hubs) but less effective for dense social clusters.
Conclusion
Bounded susceptibility demonstrates that standard epidemic models are highly sensitive to assumptions about infection rates. These results suggest that network designers and public health officials should account for physical or cognitive limits when modeling information or disease spread, as these constraints can non-intuitively alter the effectiveness of intervention strategies.
Summary updated