Coordination versus Selfishness: The Impact on Traffic Efficiency

Preprint Aug 2026

Tolls for Dynamic Equilibrium Flows

A duality-based characterization of implementability of dynamic edge flows for the multi-source, multi-destination case and a non-trivial proof that this assumption is always fulfilled for finitely supported edge flows with costs representing weighted travel times are provided.

Lukas Graf, Tobias Harks, Julian Schwarz · 0 citations

Open access Aug 2026

Effects of real-time and potential traffic congestion on network throughput

To reduce traffic congestion and improve network throughput, this study focuses on the dynamic routing process by constructing a unified path-cost function that integrates both a real-time traffic congestion index (derived from instantaneous node queue lengths) and a potential traffic congestion index (estimated from the expected number of future paths traversing each node based on active routing decisions). Through extensive simulations on Barabási–Albert scale-free network and Erdös–Rényi random network, we systematically investigate how these two types of congestion information affect network throughput. The results reveal that real-time traffic congestion is more significant than potential traffic congestion in improving network throughput. By comparison with five routing algorithms, it is found that the routing strategy considering both real-time and potential traffic congestion are more efficient. The main contribution of this work is to reveal the impacts of real-time and potential congestion on network throughput, clearly establishing their primary-secondary relationship in throughput control, and providing empirical guidelines and parameter-tuning recommendations for hybrid routing designs that combine immediate responsiveness with predictive awareness.

Kai-Tian Luo, Gang Liu · Scientific Reports · 0 citations

Open access Jul 2026

Endogenous network formation with diffusion incentives

This paper proposes a non-cooperative game of network formation in which creating access to the network – for oneself and others – is a public good. Link formation therefore reflects a trade-off between the social value of links and their private cost. The paper characterizes the architecture of strict Nash networks. Strict Nash equilibria are either flat or hierarchical. Flat equilibria consist of a wheel (i.e., a directed cycle) that may or may not encompass all agents. Hierarchical equilibria are trees in which some agents act as intermediaries between otherwise disconnected groups. When the cost of link formation is sufficiently low, the exhaustive wheel is the unique equilibrium architecture and best-response dynamics converge to it with probability one. For intermediate costs, multiple equilibrium architectures coexist and can be Pareto-ranked: the exhaustive wheels dominate the non-exhaustive ones and the empty network, and trees dominate the empty network.

Noémie Cabau · Economic Theory · 0 citations

Open access Aug 2026

Decoupling promotes cooperation in interdependent taxi-carpooling networks

Taxi carpooling is regarded as an effective strategy for alleviating urban traffic congestion, yet its actual adoption rate falls considerably short of expectations. Travelers face a social dilemma: individual rationality favors solo travel as a dominant strategy, whereas collective welfare maximization requires cooperative carpooling. Departing from traditional empirical paradigms, this study reveals the nonlinear evolutionary mechanisms of carpooling behavior from a complex network perspective. An interdependent two-layer supply–demand network model is constructed, where the passenger (demand) layer and the taxi (supply) layer are coupled through probabilistic connections. A Prisoner’s Dilemma game is embedded to simulate strategy evolution, where travelers choose between carpooling (cooperation) and non-carpooling (defection) and update strategies by imitating more successful neighbors. Extensive Monte Carlo simulations reveal an unexpected finding: lower inter-layer connection probabilities, lower connectivity degrees, and lower coupling degrees promote higher carpooling density—a network decoupling effect. When the temptation to defect exceeds a critical threshold (approximately T = 2.5), cooperation collapses regardless of network structure, revealing a nonlinear phase transition. Increased node arrival rates significantly promote cooperation in small- to medium-scale networks through continuous network renewal. These findings provide actionable policy insights for urban transportation managers while contributing to a broader understanding of cooperation dynamics on interdependent networks.

Qiang Xiao, Chunxue Yang, Kai Yan · AIP Advances · 0 citations

Open access Jul 2026

Addressing congestion in time-expanded networks: a lifeboat allocation model for maritime evacuations

This paper addresses the challenge of congestion in time-expanded networks, focusing on a case study related to maritime evacuations. The problem is made complex by an endogenous relationship between inputs and outputs, where the assignment of flow to an edge leads to increased congestion, which reflects in later arrivals and changes on the overall network topology. This dynamic interaction between flow and congestion is central to the problem, as it results in a feedback loop that complicates the identification of optimal evacuation paths. The study presents an iterative algorithm inspired by the network-simplex method, designed to handle the evolving nature of congestion while minimizing evacuation time. While the primary case study involves cruise ship evacuations, the approach is generalizable to other cases where congestion and nonlinear flow dynamics are significant factors. By considering lifeboat capacity, passenger mobility restrictions, and the impact of congestion on network structure, this work provides a practical initial plan for an evacuation off-shore, considering a congested, time-expanded network setting.

Andres Velez, Stein W. Wallace · TOP - An Official Journal of... · 0 citations

Jul 2026

The Optimization Trilemma: Efficiency, Comfort and Fairness in Decentralized Multi-agent Coordination

The problem of fair multi-agent coordination in decentralized settings is one of the most pressing challenges for building efficient collaborative systems. Resource allocation is based on optimized collective arrangements accounting for agents'needs. Such coordination should not only be computationally efficient but also account for fairness, i.e., equitable redistribution of costs incurred by all agents. Recent literature has proposed several algorithms that efficiently determine optimal plan combinations balancing system-wide efficiency and individual discomfort of agents in a centralized setting. However, these works do not address equitable resource optimization in fully decentralized scenarios, specifically, the optimized redistribution of discomfort among coordinating agents so that none experiences a discomfort level that could lead to loss of incentive or polarization that can disrupt planned operations. In this work, we study the problem of optimizing three objectives: (i) system-wide efficiency, (ii) individuals'comfort and (iii) fairness (i.e., balancing of incurred discomfort costs) in decentralized multi-agent coordination. We design a novel model to optimize those three orthogonal objectives, without any substantial increase in communication and computational overhead. Through experiments on two real-world datasets, we validate the model and demonstrate that it can achieve fairer optimization outcomes, while satisfying agents'preferences and system goals.

Jovan Nikolic, M. Zuziak, Evangelos Pournaras · arXiv.org · 0 citations