A Survey of Methods and Architectures for Traffic and Time Control Optimization Using Sensor-Driven Transmission Control Systems with MANET and Quaternion Generative Adversarial Kookaburra Optimization Networks

Abstract

Traffic congestion and inefficient signal timing remain critical challenges in modern intelligent transportation systems (ITS), especially in rapidly urbanizing environments. Recent advancements in sensor-driven transmission control systems, Mobile Ad Hoc Networks (MANET), and artificial intelligence (AI) have significantly improved traffic and time control optimization. This paper presents a comprehensive survey of methodologies and architectures integrating real-time sensing, distributed communication, and intelligent optimization techniques, including deep learning, graph neural networks (GNN), reinforcement learning (RL), and generative adversarial networks (GAN). The proposed framework emphasizes the integration of Quaternion Generative Adversarial Kookaburra Optimization Networks (QGAKON) for enhanced spatio-temporal modeling and adaptive decision-making. Sensor-based data acquisition combined with MANET enables decentralized, infrastructure-less communication, improving scalability and responsiveness. Recent studies demonstrate that hybrid AI-driven traffic control systems outperform traditional fixed-time and actuated control systems in reducing congestion, travel time, and energy consumption. This survey critically analyzes existing approaches, highlights comparative performance metrics, and identifies research gaps. The findings indicate that integrating AI, optimization algorithms, and decentralized communication frameworks can significantly enhance traffic efficiency, safety, and sustainability in smart cities.