Ultra-high reliability communications are greatly vital for massive user connections and network sensors such as Internet of Things (IoT) devices in the beyond 5th generation (B5G) communication networks. Simultaneous Wireless Information and Power Transfer (SWIPT) has evolved as an efficient strategy through prolonging the battery life by harvesting energy from radio-frequency (RF) signals. Moreover, cooperative relay sensors in IoT network can enhance the network coverage. In this paper, we investigate the multiple-input multiple-output (MIMO) SWIPT relaying systems with direct link, where the relay node splits the received RF signals into two power streams, one for information decoding (ID) and the other for energy harvesting (EH). Under per-antenna power constraints, the relay node forwards the ID signal to the destination. We propose a channel diagonalization scheme based on singular value decomposition (SVD) and eigenvalue decomposition (EVD) to design the precoding matrix of source and the amplifying matrix of relay respectively. Moreover, the closed-form expression of the optimal power splitting (PS) factor diagonal matrix is obtained to maximize the sum-rate of MIMO SWIPT relaying system. Finally, the efficiency of proposed precoding and PS factor design is verified by the presenting numerical results.

超高可靠性通信对于大规模用户连接和网络传感器（例如第五代（B5G）以后的物联网（IoT）设备）至关重要。通过从射频（RF）信号中收集能量来延长电池寿命，同步无线信息和电力传输（SWIPT）已经成为一种有效的策略。此外，物联网网络中的协作式中继传感器可以扩大网络覆盖范围。在本文中，我们研究了具有直接链接的多输入多输出（MIMO）SWIPT中继系统，其中中继节点将接收到的RF信号分为两个功率流，一个用于信息解码（ID），另一个用于能量收集（EH）。在每天线功率限制下，中继节点将ID信号转发到目的地。提出了一种基于奇异值分解（SVD）和特征值分解（EVD）的信道对角化方案，分别设计了信源的预编码矩阵和中继的放大矩阵。此外，获得最优功率分配（PS）因子对角矩阵的闭式表达式，以最大化MIMO SWIPT中继系统的总和。最后，通过数值结果验证了所提预编码和PS因子设计的效率。