Leveraging Game Theory for Intelligent and Adaptive Traffic Signal Systems

Document Type

Conference Paper

Publication Date

11-2025

Publication Title

2025 IEEE International Conference on Future Machine Learning and Data Science (FMLDS)

Abstract

Traffic congestion is a major growing problem in urban areas, where inefficient signal timing plans result in delay, emissions, and safety risks. This paper proposes an adaptive traffic signal controller for an isolated intersection utilizing connected vehicle telemetry and game-theoretical principles. This approach resulted in a fair allocation of green time to each signal phase. The proposed model was built on Python and SUMO software, and a comparison was made against a predetermined fixed signal timing plan. The model was then evaluated using (Automated Traffic Performance Measures) ATSPMs and standard intersection measures. Assess the model on an isolated intersection in Columbus, Ohio, and the model showed a significant performance increase. Results showed an improvement of 29.62% in total delay, 28.69% in average speed, 5% in travel time, 15% average queue length, 19.83% in average fuel consumption, 19.83% in carbon monoxide emissions, 45.8% in split failures, and 4% in arrivals on red relative to the baseline model.

Comments

Paper presented at 2025 IEEE International Conference on Future Machine Learning and Data Science (FMLDS) held in Los Angeles, CA, November 2-5, 2025

Cleveland State University for funding this study through its Faculty Research and Development (FRD) program and the Rural Safe, Efficient, and Advanced Transportation (R-SEAT) Center, a Tier-1 University Transportation Center (UTC) funded by the United States Department of Transportation (USDOT), through the agreement number 69A3552348321.

DOI

10.1109/FMLDS67896.2025.00093

Link to Full Text

Share

COinS