In this letter, we consider the dual-hop and three-hop multiple-input multiple-output (MIMO) sensor systems using multiple Khatri-Rao space-time (M-KRST) coding at the unmanned aerial vehicles and sensor nodes. According to the tensor structure of the communication links, the signals received at the fusion center form two separate nested tensors, which satisfy a nested parallel factor (PARAFAC) model and a nested parallel Tucker2 (PARATUCK2) model. Exploiting this nested structure, two semi-blind receivers based on an alternating least squares algorithm are formulated for a joint estimation of symbols and channels. Compared with the two-stage training method, simulation results demonstrate the superiority of the proposed semi-blind receivers in terms of normalized mean square error and bit error rate.

中文翻译：

---

基于嵌套张量模型的无人机 M-KRST 编码 MIMO 系统半盲接收机


在这封信中，我们考虑在无人机和传感器节点上使用多个 Khatri-Rao 空时 (M-KRST) 编码的双跳和三跳多输入多输出 (MIMO) 传感器系统。根据通信链路的张量结构，融合中心接收到的信号形成两个独立的嵌套张量，满足嵌套并行因子（PARAFAC）模型和嵌套并行Tucker2（PARATUCK2）模型。利用这种嵌套结构，两个基于交替最小二乘算法的半盲接收器被制定用于符号和信道的联合估计。与两阶段训练方法相比，仿真结果证明了所提出的半盲接收器在归一化均方误差和误码率方面的优越性。