Simultaneously transmitting and reflecting reconfigurable intelligent surface (STAR-RIS) is a novel technology which enables the full-space coverage by splitting the incident signal into reflected and transmitted signals. In this letter, a multi STAR-RIS-aided system using non-orthogonal multiple access (NOMA) in an uplink transmission is considered, where the multi-order reflections among multiple STAR-RISs assist the transmission from the single-antenna users to the multi-antenna base station (BS). Specifically, the total sum rate maximization problem is solved by jointly optimizing the active beamforming, power allocation, transmission and reflection beamforming at the STAR-RIS, and user-STAR-RIS association indicator. To solve the non-convex optimization problem, a novel deep reinforcement learning algorithm is proposed which is the combination of meta-learning and deep deterministic policy gradient (DDPG), namely Meta-DDPG. Numerical results demonstrate that the proposed Meta-DDPG algorithm outperforms the conventional DDPG algorithm.

翻译：同时发射与反射可重构智能表面（STAR-RIS）是一种通过将入射信号分割为反射信号和透射信号以实现全空间覆盖的新型技术。本文考虑在上行传输中采用非正交多址接入（NOMA）的多 STAR-RIS 辅助系统，其中多个 STAR-RIS 间的多阶反射辅助单天线用户向多天线基站（BS）的传输。具体而言，通过联合优化主动波束赋形、功率分配、STAR-RIS 处的发射与反射波束赋形以及用户-STAR-RIS 关联指标，解决了总速率最大化问题。为解决该非凸优化问题，本文提出了一种融合元学习与深度确定性策略梯度（DDPG）的新型深度强化学习算法，即 Meta-DDPG。数值结果表明，所提 Meta-DDPG 算法优于传统 DDPG 算法。