Internet of Things (IoT) is a promising paradigm to connect massive number of devices in future wireless communications while satisfying various quality of service (QoS) requirements. In this paper, we consider a QoS-constrained IoT system operating with finite blocklength (FBL) codes to support low latency communications. Two data arrival models are considered, namely, constant-rate arrival and ON-OFF discrete-time Markov arrival. The throughput performance is studied for both arrival models under statistical queuing constraints and deadline limits. For the scenario with instantaneous channel state information (CSI), we derive the QoS-constrained throughput expressions for both arrival models. Subsequently, an instantaneous-CSI-driven optimal power control algorithm is proposed to maximize the throughput, while guaranteeing a certain reliability target. In addition, we consider a scenario with only average CSI being available at the transmitter and propose to apply hybrid automatic repeat request (HARQ) schemes to improve the FBL performance. The decoding error probability and the outage probability are first characterized, following which the distribution of transmission period is derived. Furthermore, the throughput expressions are provided for both types of arrivals. Via numerical analysis, the impact of error probability, fixed transmission rate, coding blocklength, and QoS constraints on the throughput is studied.

中文翻译：

---

具有 QoS 约束和有限块长度代码的低延迟物联网系统的吞吐量分析

物联网 (IoT) 是一种很有前途的范例，可以在未来的无线通信中连接大量设备，同时满足各种服务质量 (QoS) 要求。在本文中，我们考虑使用有限块长度 (FBL) 代码运行的 QoS 受限物联网系统，以支持低延迟通信。考虑了两种数据到达模型，即恒定速率到达和 ON-OFF 离散时间马尔可夫到达。在统计排队约束和期限限制下研究了两种到达模型的吞吐量性能。对于具有瞬时信道状态信息 (CSI) 的场景，我们推导出两种到达模型的 QoS 约束吞吐量表达式。随后，提出了一种瞬时 CSI 驱动的最优功率控制算法来最大化吞吐量，同时保证一定的可靠性目标。此外，我们考虑在发射机处只有平均 CSI 可用的情况，并建议应用混合自动重传请求 (HARQ) 方案来提高 FBL 性能。首先表征解码错误概率和中断概率，然后导出传输周期的分布。此外，还为两种类型的到达提供了吞吐量表达式。通过数值分析，研究了错误概率、固定传输速率、编码块长度和QoS约束对吞吐量的影响。首先表征解码错误概率和中断概率，然后导出传输周期的分布。此外，还为两种类型的到达提供了吞吐量表达式。通过数值分析，研究了错误概率、固定传输速率、编码块长度和QoS约束对吞吐量的影响。首先表征解码错误概率和中断概率，然后导出传输周期的分布。此外，还为两种类型的到达提供了吞吐量表达式。通过数值分析，研究了错误概率、固定传输速率、编码块长度和QoS约束对吞吐量的影响。