In this paper, we study the resource allocation design for non-orthogonal multiple access (NOMA)-based cellular massive Internet-of-Things (IoT) enabled with simultaneous wireless information and power transfer (SWIPT). The design is formulated as a non-convex optimization problem, which takes into account practical and adverse factors, e.g., the channel uncertainty during channel state information (CSI) acquisition, the non-linear receiver during energy harvesting (EH) and the imperfect successive information cancellation (SIC) during the information decoding (ID). The originally harmful co-channel interference in massive access is coordinated to strike a balance between efficient information transmission and efficient energy harvesting via spatial beamforming. Subsequently, two robust beamforming algorithms are designed from the aspects of the weighted sum rate maximization and the total power consumption minimization, respectively. It is found that overall performance can be improved by adding BS antennas due to more array gains. Moreover, it is proved that the proposed algorithms can effectively alleviate the influence of adverse practical conditions and achieve the best performance compared to the baseline ones, which demonstrates the effectiveness and robustness of proposed algorithms for cellular massive IoT.

中文翻译：

---

具有 SWIPT 的基于 NOMA 的蜂窝大规模物联网的稳健波束成形

在本文中，我们研究了基于非正交多址 (NOMA) 的蜂窝式大规模物联网 (IoT) 的资源分配设计，支持同时无线信息和功率传输 (SWIPT)。该设计被表述为一个非凸优化问题，它考虑了实际和不利因素，例如信道状态信息 (CSI) 获取期间的信道不确定性、能量收集 (EH) 期间的非线性接收器和不完美的连续信息解码 (ID) 期间的信息消除 (SIC)。协调大规模接入中原本有害的同信道干扰，以通过空间波束成形在高效信息传输和高效能量收集之间取得平衡。随后，分别从加权和速率最大化和总功耗最小化两个方面设计了两种鲁棒的波束成形算法。发现由于更多的阵列增益，可以通过添加 BS 天线来提高整体性能。此外，证明了所提出的算法可以有效减轻不利实际条件的影响，并与基线算法相比达到最佳性能，证明了所提出算法在蜂窝大规模物联网中的有效性和鲁棒性。