Recent developments in industrial wireless sensor networks (IWSNs) have revolutionized industrial automation systems. However, harsh industrial environment poses great challenges to a time-critical and reliable wireless communication. For instance, effects of multipath fading, noise and co-channel interference can have unpredictable and time-varying impacts on the propagation channel, leading to the failure of on-time packet delivery. To address this problem, in this paper we propose a channel-based Optimal Back-off Delay Control (OBDC) scheme which can minimize the total time a packet spends in the sensor node (TSN) by assessing the features of a generic wireless channel. Specifically, we first explore the channel impairments by investigating the probability density function (PDF) of the level crossing rate (LCR) of the received signal in the industrial wireless environment. Then, with the obtained channel assessment results, we develop a phase-type semi-Markov model to investigate the probability distribution of the back-off delay of a packet in the sensor node (SN). The probability distribution of the back-off delay can be further substituted with TSN according to the queuing theory. The proposed OBDC scheme examines the Kullback-Leibler (KL) divergence between the obtained distribution of TSN and the packet arrival rate, and reduces the TSN according to an objective function which is constantly renewed in every transmission round with regard to a delay constraint. The simulation results show that the OBDC scheme can reduce TSN and guarantee to keep the TSN in an acceptable range even though the wireless channel is impaired by interference effects. It also shows that the OBDC scheme can reduce the proportion of packets meeting their deadline to the total packets in transmission when the number of SN and LCR changes

中文翻译：

---

延迟受限工业无线传感器网络中基于信道的最优回退延迟控制

工业无线传感器网络 (IWSN) 的最新发展已经彻底改变了工业自动化系统。然而，恶劣的工业环境对时间紧迫且可靠的无线通信提出了巨大挑战。例如，多径衰落、噪声和同信道干扰的影响会对传播信道产生不可预测且随时间变化的影响，从而导致无法按时交付数据包。为了解决这个问题，在本文中，我们提出了一种基于信道的最优回退延迟控制 (OBDC) 方案，该方案可以通过评估通用无线信道的特征来最小化数据包在传感器节点 (TSN) 中花费的总时间。具体来说，我们首先通过研究工业无线环境中接收信号的电平交叉率 (LCR) 的概率密度函数 (PDF) 来探索信道损伤。然后，根据获得的信道评估结果，我们开发了一个相位型半马尔可夫模型来研究传感器节点 (SN) 中数据包回退延迟的概率分布。根据排队论，回退延迟的概率分布可以进一步用TSN代替。所提出的 OBDC 方案检查所获得的 TSN 分布与数据包到达率之间的 Kullback-Leibler (KL) 散度，并根据目标函数减少 TSN，该目标函数在每个传输轮中不断更新，并考虑延迟约束。仿真结果表明，即使无线信道受到干扰影响，OBDC方案也能降低TSN并保证将TSN保持在可接受的范围内。这也表明当SN和LCR的数量发生变化时，OBDC方案可以减少满足其截止时间的数据包占传输数据包总数的比例