TRUAV applies distributed multi-agent RL to joint UAV trajectory planning and routing in IoT-enabled vehicular networks

arXiv·low signalXBlueskyLinkedInCopy link

Posted July 27, TRUAV treats UAV placement and network routing as a single distributed multi-agent reinforcement learning problem rather than optimizing them separately, targeting UAV-aided IoT vehicular ad-hoc networks where topology changes faster than centralized planning can react. The framing is the notable part for agent builders: coupled control and communication decisions handled by cooperating agents with local observations, instead of a central orchestrator. Single-source and early-stage — treat as a research signal on decentralized multi-agent coordination, not a production pattern.