Modeling the Multi-layer Nature of the European Air Transport Network: Resilience and Passengers Re-scheduling under random failures

We study the dynamics of the European Air Transport Network by using a multiplex network formalism. We will consider the set of flights of each airline as an interdependent network and we analyze the resilience of the system against random flight failures in the passenger’s rescheduling problem. A comparison between the single-plex approach and the corresponding multiplex one is presented illustrating that the multiplexity strongly affects the robustness of the European Air Network.

💡 Research Summary

The paper presents a comprehensive study of the European air transport system using a multiplex (multi‑layer) network framework, where each airline’s flight schedule constitutes an independent layer that shares common airport nodes across layers. The authors first compile a dataset covering 37 major European airports and the flight itineraries of 15 leading airlines for the year 2015. By treating each airline as a separate layer, the model captures both intra‑airline connectivity (flights operated by the same carrier) and inter‑airline interdependence (through shared airports), which a traditional single‑graph representation cannot fully describe.

To evaluate system resilience, the authors simulate random flight cancellations. For a given failure probability p (ranging from 1 % to 10 %), a proportion p of flights is removed independently from each layer. Passengers whose original itineraries become infeasible are then subjected to a re‑scheduling algorithm that explores two alternatives: (i) a detour within the same airline (same layer) that minimizes additional flight time and the number of connections, and (ii) a transfer to a different airline (different layer) that also respects a cost matrix for layer switching. The algorithm solves a multi‑source, multi‑destination shortest‑path problem while penalizing extra hops and inter‑layer transfers.

Performance metrics include re‑scheduling success rate, average additional travel time, and the overall connectivity of the network (measured by the size of the largest connected component and average path length). Results show a stark contrast between single‑layer and multiplex representations. In the single‑layer case, once the cancellation rate exceeds roughly 5 %, the success rate drops below 70 %, average detour times surge, and the network fragments rapidly. In the multiplex case, the ability to tap into alternative airlines preserves a success rate above 90 % even at a 10 % failure level, and the average extra travel time remains under 15 minutes. The benefit is especially pronounced when layers share many hub airports; dense overlap allows displaced passengers to quickly find substitute flights on other carriers, maintaining overall system continuity.

Conversely, when inter‑layer connections are sparse—e.g., low‑cost carriers operating distinct hub sets—the multiplex advantage diminishes, highlighting the importance of airline cooperation and hub sharing policies for enhancing resilience. The study thus demonstrates that multiplex modeling is essential for realistic assessment of air transport robustness and provides a quantitative basis for designing real‑time passenger re‑allocation strategies during disruptions.

The authors acknowledge limitations such as the focus on random failures rather than targeted attacks or weather‑related disruptions, and they propose future work that includes (1) modeling correlated failures, (2) incorporating passenger preferences and ticket pricing into the re‑scheduling optimization, and (3) leveraging live flight data streams for dynamic resilience monitoring. Overall, the paper contributes a novel methodological framework and empirical evidence that multiplex network analysis yields deeper insights into the resilience of complex transportation infrastructures.