Meta-Learning-Enhanced Task Assignment and Resource Scheduling for UAV-Assisted WSNs in 6G-Enabled ITS

The integration of unmanned aerial vehicles (UAVs), wireless sensor networks (WSNs), and 6 G technologies is transforming the design of next-generation Intelligent Transportation Systems (ITS), enabling real-time, energy-efficient, and adaptive urban mobility services. This paper proposes a novel meta-learning-enhanced UAV-assisted WSN architecture, tailored for large-scale, dynamic ITS environments characterized by high vehicular mobility, dense sensor deployments, and stringent latency requirements. We introduce a new task assignment framework to address critical challenges such as coverage limitations at urban intersections, energy constraints of sensor nodes, and the need for rapid decision-making under complex traffic conditions. We aim to maximize energy-efficient data throughput (EEDT) and ensuring Quality of Service (QoS) in UAV-assisted WSNs. The framework jointly optimizes traffic sensor node selection, UAV trajectory planning, and communication resource allocation by formulating the problem as a Constrained Markov Decision Process (CMDP). To solve this, we develop the Meta-Learning Weighted Multi-Agent Deep Deterministic Policy Gradient (MW-MAD3PG) algorithm, which embeds model-agnostic meta-learning (MAML) within a cooperative multi-agent reinforcement learning (MADRL) setting. MW-MAD3PG enables UAV agents to rapidly adapt to dynamic traffic patterns, fluctuating road conditions, and evolving network states with minimal retraining, ensuring high energy efficiency and system scalability. Extensive simulations show that the proposed framework achieves up to a 25% improvement in UAV coordination efficiency, a 30% increase in data offloading capacity, and substantial enhancements in energy-aware operations compared to baseline methods such as MADDPG, Meta-SGD, and Meta-Q-Learning. These results validate our architecture and framework as robust, intelligent solutions for future 6G-enabled ITS, supporting resilient, real-time, and energy-optimized traffic monitoring and control.