Autonomous traffic management in smart cities presents a significant challenge due to the growing number of vehicles and the complexity of urban environments. Traditional traffic control systems lack the real-time adaptability to manage dynamic traffic patterns, leading to congestion, accidents, and inefficient energy usage. This research explores the integration of IoT networks to create a decentralized, real-time traffic management system capable of autonomous decision-making. By utilizing Vehicle-to-Infrastructure (V2I) communication, sensors, and edge computing, the proposed system monitors traffic conditions, predicts congestion points, and dynamically adjusts traffic signals and rerouting strategies. A machine learning model analyzes traffic patterns and optimizes flow in real-time, improving overall system efficiency. This research demonstrates that IoT-based traffic management systems offer substantial improvements over traditional methods, setting the foundation for future autonomous traffic control solutions in increasingly connected urban environments.