DRL Framework for Dynamic HAPS Positioning in Wind-Disturbed Maritime Networks
A deep reinforcement learning framework using Proximal Policy Optimization (PPO) is proposed for dynamic positioning of High-Altitude Platform Station (HAPS) base stations in maritime networks affected by stratospheric wind disturbances. The system employs a centralized DRL agent on a coordinator HAPS to control multiple serving HAPS, utilizing radio measurements and network feedback to adapt to ship mobility and wind effects. Simulations demonstrate improved coverage stability and system throughput under wind disturbances. The research addresses challenges in providing wireless coverage to maritime regions lacking terrestrial infrastructure.
Key facts
- Proposes a DRL-based framework for dynamic positioning of HAPS base stations.
- Uses Proximal Policy Optimization (PPO) algorithm.
- Centralized DRL agent deployed on a coordinator HAPS controls multiple serving HAPS.
- Addresses stratospheric wind disturbances and ship mobility.
- Simulations show improved coverage stability and throughput.
- Aims to provide wireless coverage in maritime regions without terrestrial infrastructure.
- Published on arXiv with ID 2605.05240v1.
- Source: arXiv preprint.
Entities
Institutions
- arXiv