I. Introduction
Emerging applications of the Industrial Internet of Things (IIoT) systems, e.g., factory automation, smart grid, vehicle networks, robotic surgeries, and tactile Internet, pursue ultrareliability, low-latency, high-data rates, high-information timeliness, etc. [1], [2], [3], [4], [5], [6]. In particular, most mission-critical applications often have stringent requirements on the reliability, latency, and throughput [7]. While the majority of the industries are in the remote areas, the capability of distant ultra-Reliable and Low-Latency Communication (uRLLC) has become one of the key performance metrics of the emerging applications in IIoT. Ensuring the conflicting Quality-of-Service (QoS) requirements in distant uRLLC is crucial and really challenging. In effect, relaying provides intuitive solution for improving the communication distance. Recent work [8] found that deploying relay in wireless networks helps improving the achievable rate of uRLLC. It is noteworthy that with half-duplex relaying, the resource utilization efficiency is low and an extra transmission delay is introduced. Accordingly, full-duplex relaying is considered to enhance the resource utilization efficiency and to avoid the extra transmission delay [9], [10], [11]. However, transmitting and receiving data at the same time would cause self-interference, which impairs the performance of the relaying and increases the error probability. Moreover, to guarantee the low-latency requirement, a short frame structure is adopted in the 5th generation (5G) new radio (NR), where the coding blocklength is short [8]. The short frame structure should also be employed in the distant uRLLC IIoT systems. Since Shannon capacity is the maximum achievable rate when the coding blocklength goes to infinity, it is no longer applicable in the finite blocklength (FBL) regime [12], [13], [14]. The FBL theory should be taken into account.