# Decentralized Cycle-Free Game-Theoretic Adaptive Traffic Signal Control: Model Enhancement and Testing on Isolated Signalized Intersections

## TL;DR

This paper improves a decentralized traffic signal control system using game theory, showing it reduces delays and queues better than traditional and AI-based methods.

## Contribution

A decentralized game-theoretic traffic signal controller is enhanced and tested, offering real-time adaptability without pre-training.

## Key findings

- The DNB controller reduced average vehicle delay by up to 54% compared to fixed-time control.
- Queue sizes were reduced by up to 63% compared to traditional methods.
- The DNB controller outperformed reinforcement learning approaches without requiring pre-training.

## Abstract

This research enhances and evaluates the performance of a Decentralized Nash Bargaining (DNB) adaptive traffic signal controller that operates a flexible National Electrical Manufacturers Association (NEMA) phasing and timing scheme responding dynamically to fluctuating traffic demands. The DNB controller is enhanced to (1) use traffic density estimates instead of queues to optimize signal timings; (2) to consider the eight-phase two-ring NEMA controller configuration within the game-theoretic approach; and (3) to consider dynamically adaptable control time steps. The enhanced DNB controller is benchmarked against (1) a fixed-time traffic signal control using the state-of-practice Webster’s method and an emerging Laguna-Du-Rakha (LDR) method for computing the optimum cycle length; (2) a state-of-the-practice actuated traffic signal control; and (3) a state-of-the-art reinforcement learning (RL) traffic signal controller presented in the literature. The controller is tested on two isolated signalized intersections, demonstrating enhanced overall intersection performance compared to the baseline pretimed and actuated controllers at various demand levels, and offers better performance than a previously developed RL controller. Specifically, the DNB controller results in a decrease in the average vehicle delay and queue size by up to 54% and 63%, respectively, compared to Webster’s state-of-the-practice pretimed control. Unlike the RL controller, the DNB controller requires no pre-training while adapting to fluctuating traffic conditions, thereby providing a flexible framework for reducing traffic congestion at signalized intersections. As such, this research contributes to the development of smarter and more responsive urban traffic control systems.