Both non-orthogonal multiple access (NOMA) and massive multiple-input multiple-output (MIMO) are the promising techniques to satisfy the foreseeable massive connectivity requirements. In this letter, a novel semi-blind channel estimation method is proposed for multi-cell massive MIMO-NOMA systems, in which a group successive interference cancellation (GSIC) assisted method is developed to mitigate pilot contamination. Specifically, the users in each cell are divided into multiple groups and the same pilot sets are reused by the different groups in order to reduce pilot overhead. Besides, by following the NOMA principle, a GSIC assisted method is proposed to eliminate inter-group pilot contamination. In simulations, it is shown that the proposed channel estimation method is capable of reducing the overlapping probability of different groups as the number of receiving antennas increases. In addition, the proposed scheme outperforms the conventional channel estimation methods in terms of normalized mean square error (NMSE) performance.

中文翻译：

---

多小区大规模 MIMO-NOMA 系统中的群连续干扰消除辅助半盲信道估计

非正交多址（NOMA）和大规模多输入多输出（MIMO）都是满足可预见的大规模连接需求的有前途的技术。在这封信中，针对多小区大规模 MIMO-NOMA 系统提出了一种新的半盲信道估计方法，其中开发了一种组连续干扰消除 (GSIC) 辅助方法来减轻导频污染。具体地，每个小区的用户被分成多个组，相同的导频集被不同的组重用，以减少导频开销。此外，通过遵循NOMA原理，提出了一种GSIC辅助方法来消除组间飞行员污染。在模拟中，结果表明，随着接收天线数量的增加，所提出的信道估计方法能够降低不同组的重叠概率。此外，所提出的方案在归一化均方误差 (NMSE) 性能方面优于传统的信道估计方法。