Ruiqi Wang

Ruiqi Wang

Joint Optimization of Multi-UAV Beamforming and Path Planning

  • Description: This project develops a learning-based framework for multi-UAV downlink communications. It decouples short-term beamforming and long-term cooperative trajectory planning, combines topology-aware GNN beamforming with MAPPO-based path planning, and follows the centralized-training decentralized-execution paradigm.
  • Role: Independent researcher, algorithm designer.
  • Keywords: GNN, MAPPO, Dec-POMDP, multi-agent reinforcement learning, wireless communications.
Two-layer learning framework for GNN beamforming and MAPPO trajectory optimization
Two-layer learning framework combining GNN-based beamforming and MAPPO-based trajectory optimization.