Research story · Queueing theory & game theory

Passenger–taxi matching with strategic arrivals

What changes when both congestion and matching matter, agents choose whether to participate, and matching itself takes time?

Queueing Systems · 2022Double-ended queuesEquilibrium & optimal policy

The problem

Passenger–taxi systems are not ordinary one-sided queues. Two populations arrive, their imbalance creates waiting on one side, and the system state affects whether a newly arriving agent finds participation worthwhile. Multiple access points and nonzero matching times make the state and decision structure richer.

Key idea

The model treats participation as a strategic decision embedded in a stochastic matching system. Queueing analysis describes the evolving state; game-theoretic analysis characterizes how rational agents respond to what they observe.

A service system can change its own demand because users respond to the congestion it creates.

Main contribution

The paper characterizes strategic behavior and studies system-level policies in a passenger–taxi double-ended queue with multiple access points and nonzero matching times.

Why it matters

The broader lesson applies to platforms and matching services: adding capacity or changing access conditions can alter behavior, so operational design should account for equilibrium response rather than treating arrivals as fixed.