Reliability in Post-Disaster Networks: A Novel Interference-Mitigation Strategy

Maurilio Matracia, Mustafa A. Kishk, Mohamed‐Slim AlouiniIEEE Open Journal of Vehicular Technology 2024 · DOI

This paper, authored by researchers from KAUST and Maynooth University, explores a counter-intuitive but effective strategy for restoring communication in the wake of a disaster: intentionally turning off surviving base stations to improve the overall reliability of the network.

TL;DR

The researchers propose a "Base Station Silencing" strategy that improves network coverage and reliability in disaster zones by strategically deactivating a ring of surviving base stations to mitigate aggregate interference. Using a novel stochastic geometry framework, they demonstrate that this approach can significantly boost connectivity for users at the heart of a calamity without requiring expensive new hardware deployments.

What Problem Does This Paper Address?

When a disaster strikes—be it an earthquake, flood, or conflict—terrestrial telecommunications infrastructure is often the first thing to fail. Towers lose power or are physically destroyed, creating "coverage holes."

While much research focuses on deploying "ad-hoc" solutions like drones (UAVs) or satellite links, these take time and money to deploy. This paper looks at the surviving infrastructure. The problem isn't just that towers are missing; it’s that the surviving towers often create a chaotic interference environment. For a user equipment (UE) near a disaster epicenter, the signal from a distant surviving tower is weak, and it is easily drowned out by the "noise" (interference) of other surviving towers further away. The authors aim to find a way to clear this "radio smog" to ensure first responders and victims can actually connect.

Methods

The authors employ Stochastic Geometry (SG), a powerful mathematical tool for modeling random spatial patterns.

  • Inhomogeneous Poisson Point Process (IPPP): They model the surviving base stations (BSs) not as a uniform grid, but as a random process where the density of stations drops significantly within the disaster zone (defined by a "Quality of Resilience" parameter, $\chi$).
  • The Silencing Ring: The core novelty is the "silencing strategy." They define a ring-shaped area ($A_r$) around the disaster zone. All functioning BSs within this ring are put into stand-by mode.
  • SINR Meta Distribution: Instead of just looking at "average coverage," which can hide the fact that many users still have zero connection, they use the meta distribution of the Signal-to-Interference-plus-Noise Ratio (SINR). This provides a granular look at the "reliability of the reliability"—essentially asking, "What percentage of users can achieve a stable connection 90% of the time?"
  • Dominant-Interferer Approximation: Mathematically, the paper provides the first derivation of a dominant-interferer-based approximation for the SINR meta distribution in an IPPP-modeled network.

Datasets

The paper is theoretical and analytical in nature. It does not use real-world datasets. Instead, the authors use Monte Carlo simulations to validate their derived mathematical expressions. They define standard system parameters (e.g., BS density of 2 BSs/km², transmit power of 20W) to create a synthetic but representative environment for testing.

Key Insights and Ideas

  • Less is More: In interference-limited environments (typical of dense cellular networks), deactivating infrastructure can actually increase the "Quality of Service" for high-priority users by clearing the path for the desired signal.
  • The Epicenter Effect: For small disasters ($r_d = 500m$), the user's exact location is less critical than the silencing strategy itself. However, for larger disasters ($r_d = 2km$), the epicenter becomes a "dead zone" for real-time traffic unless silencing is implemented aggressively.
  • The Reliability Gap: The authors found that even if a network isn't resilient at all ($\chi \to 0$), silencing can preserve a basic level of reliability that would otherwise be completely lost.
  • Optimization Trade-off: There is a "cost" to silencing—users living inside the silencing ring lose their local towers. The paper provides an optimization framework to maximize coverage in the disaster zone while ensuring the "cost" to the surrounding ring doesn't exceed a specific threshold (the "average coverage priority ratio").

Key Results

  • Coverage Boost: In a small disaster with zero resilience, optimizing the silencing radii can boost average coverage from 67.4% to 83.1%.
  • Reliability Gains: In partially resilient networks, the researchers showed that with optimal silencing, roughly 97% of users could achieve a reliability threshold of 0.6, compared to only 78% without silencing.
  • Distance Sensitivity: The benefit of silencing is most pronounced for users at the very center of the disaster. As the user moves toward the edge of the disaster zone, the benefit of silencing decreases because they are closer to surviving (and now serving) base stations.

Limitations and Future Work

  • Network Capacity: The model primarily addresses signal quality (SINR). It does not account for capacity—if you turn off half the towers, the remaining towers might become overloaded by too many users trying to connect at once.
  • Single-Tier Model: The study assumes a single tier of macro base stations. Real-world "HetNets" (Heterogeneous Networks) involve macro, micro, and femto cells, which would complicate the interference and silencing patterns.
  • Static Nature: The model is a snapshot in time. Future work could look at dynamic silencing that shifts as first responders move through a zone.
  • Future Directions: The authors suggest integrating this with load analysis and exploring spectrum allocation schemes that could mitigate interference without needing to turn BSs off entirely.

Summary updated

Abstract

We hereby present a novel interference mitigation strategy specifically designed to enhance the quality of service that a typical terrestrial user equipment (UE) would experience after the occurrence of a calamity. In particular, we devise a novel stochastic geometry framework where the functioning ground base stations are modeled as an inhomogeneous Poisson point process, and promote proper silencing as an effective solution to improve both coverage and reliability (which is usually overlooked in emergency scenarios); in particular, the latter is evaluated by means of the signal-to-interference-plus-noise ratio (SINR) meta distribution performance metric. The derived downlink performances assume Rayleigh fading conditions for all wireless links. The numerical results show insightful trends in terms of both average coverage probability (which is optimized by choosing the best area for applying the silencing strategy) and SINR meta distribution, depending on: distance of the UE from the disaster epicenter (henceforth intended as the center of the area where the BS can be damaged), disaster radius (also referring to the latter area), and quality of resilience of the terrestrial network. The aim of this paper is therefore to prove the effectiveness of proper silencing in emergency scenarios, at least from the coverage and reliability perspectives.
BibTeX
@article{matracia2024reliability,
  author = {Maurilio Matracia and Mustafa A. Kishk and Mohamed‐Slim Alouini},
  title = {{Reliability in Post-Disaster Networks: A Novel Interference-Mitigation Strategy}},
  journal = {IEEE Open Journal of Vehicular Technology},
  year = {2024},
  doi = {10.1109/ojvt.2024.3353611},
}
Figure