Role of layer positioning in multilayer traffic networks: A trade-off between path efficiency and load balancing.

Abstract

Understanding how structural organization shapes dynamical performance is essential for designing efficient multilayer infrastructures. As modern transportation systems become increasingly multimodal, integrating layers with distinct connectivity patterns raises a fundamental yet underexplored question: How should structurally heterogeneous layers be arranged to maximize global traffic capacity? To address this, we develop a multilayer framework with heterogeneous intralayer connectivity and examine how different choices of which layer occupies the structural core affect overall transport efficiency. A notable phenomenon emerges: placing the structurally strongest layer at the core does not always deliver the best performance. Although this arrangement enhances path efficiency when within-layer traffic dominates, increasing cross-layer flow causes the strongly connected core to become the first bottleneck, sharply reducing capacity. Beyond a critical regime, positioning the weakest layer instead at the core becomes the superior strategy, as it naturally disperses load across layers and mitigates congestion. This performance reversal arises from a fundamental trade-off between path efficiency and load balancing, governed by bottleneck migration across layers. A theoretical framework based on effective betweenness accurately predicts these transitions and matches extensive simulations. Importantly, the positioning effect remains robust across varying degrees of heterogeneity, network scales, and representative network topologies. These results establish layer positioning as a universal and tunable structural mechanism, offering guidance for the design of multilayer communication and transportation systems.