A Distributed Congestion-Aware Traffic Light Signal Controller for Sustainable Urban Mobility

Abstract

Improving urban mobility sustainability requires solutions that reduce emissions while remaining compatible with existing infrastructures. This paper presents a distributed congestion-aware traffic light controller that relies exclusively on locally observable traffic information. The approach extends a rule-based adaptive baseline through a cost function that incorporates upstream demand, downstream occupancy, and local saturation. An adaptive mechanism dynamically balances speed-oriented and congestion-oriented logic based on real-time intersection occupancy. Evaluated through microscopic simulation on the Bologna urban network and a Manhattan-style grid, the controller improves environmental and mobility metrics under medium-to-high loads. Crucially, results show that selective deployment on a subset of critical intersections can provide greater benefits than full-network deployment in realistic scenarios, offering a cost-effective pathway for sustainable smart-city transitions.