Scholarlink

In this chapter, we extend the hierarchical DRL framework to AoI minimization problem in an IRS-assisted wireless network. Considering dynamic channel conditions, the outer-loop PPO is employed to adapt the per-slot scheduling decision for each sensor device to upload sensing information to an AP. If no sensor device is scheduled due to energy shortage or deteriorating channel conditions, the AP will schedule downlink wireless power transfer to all sensor devices with energy harvesting capabilities. Given the outer-loop scheduling decision generated by PPO, the inner-loop optimization refines the transmission control parameters by simple relaxation and approximations. Such a hierarchical PPO is shown to achieve significantly higher cumulative rewards and faster convergence compared to model-free DRL approaches.

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Hierarchical DRL for IRS-Assisted AoI Minimization

  • 章节

  • 发表日期：

    2026年

查看全文
  • Shimin Gong,
  • Dusit Niyato,
  • Bo Gu,
  • Kaibin Huang

摘要

In this chapter, we extend the hierarchical DRL framework to AoI minimization problem in an IRS-assisted wireless network. Considering dynamic channel conditions, the outer-loop PPO is employed to adapt the per-slot scheduling decision for each sensor device to upload sensing information to an AP. If no sensor device is scheduled due to energy shortage or deteriorating channel conditions, the AP will schedule downlink wireless power transfer to all sensor devices with energy harvesting capabilities. Given the outer-loop scheduling decision generated by PPO, the inner-loop optimization refines the transmission control parameters by simple relaxation and approximations. Such a hierarchical PPO is shown to achieve significantly higher cumulative rewards and faster convergence compared to model-free DRL approaches.