Merging lanes—fairness through communication

• The formal definition of a model describing vehicle behavior when two lanes merge.
• The specification of a fairness criterion for this behavior.
• A distributed algorithm that results in good fairness under real-world conditions.
• Simulation results analyzing the impact of participating vehicles.

The merging of two lanes is a common traffic scenario. In this paper we derive a formal model for the behavior of vehicles in this scenario. We discuss the question of how fairness of a merging process can be defined and introduce the notion of free-flow fairness. We first show how optimal fairness could be achieved if all vehicles were omniscient and willing to follow a given strategy. We then move to a more realistic setting, where only a subset of vehicles participates in our merging scheme and where wireless communication is limited and unreliable. By means of analysis and simulation we show that a simple beacon-based approach yields very good fairness even if only 1% of the vehicles participate.