Yu Yun-Shuai, Chen Yeong-Sheng
Department of Computer Science and Information Engineering, National Formosa University, Yunlin 632301, Taiwan.
Department of Computer Science, National Taipei University of Education, Taipei 106320, Taiwan.
Sensors (Basel). 2020 Jul 17;20(14):3978. doi: 10.3390/s20143978.
Industrial wireless sensor networks (IWSNs) are a key technology for smart manufacturing. To identify the performance bottlenecks in an IWSN before its real-world deployment, the IWSN must first be evaluated through simulations using an error model which accurately characterizes the wireless links in the industrial scenario within which it will be deployed. However, the traditional error models used in most IWSN simulators are not derived from the real traces observed in industrial environments. Accordingly, this study first measured the transmission quality of IEEE 802.15.4 in a one-day experiment in a manufacturing factory and then used the measurement records to construct a second-order Markov frame-level error model for simulating the performance of an IWSN. The proposed model was incorporated into the simulator of OpenWSN, which is an industrial WSN implementing the related IEEE and IETF standards. The simulation results showed that the proposed error model improved the accuracy of the estimated transmission reliability by up to 12% compared to the original error model. Moreover, the estimation accuracy improved with increasing burst losses.
工业无线传感器网络（IWSN）是智能制造的关键技术。为了在实际部署之前识别IWSN中的性能瓶颈，必须首先使用误差模型通过仿真对IWSN进行评估，该误差模型要能准确表征其将被部署的工业场景中的无线链路。然而，大多数IWSN模拟器中使用的传统误差模型并非源自工业环境中观察到的实际踪迹。因此，本研究首先在一家制造工厂进行了为期一天的实验，测量了IEEE 802.15.4的传输质量，然后使用测量记录构建了一个二阶马尔可夫帧级误差模型，用于模拟IWSN的性能。所提出的模型被纳入OpenWSN模拟器，OpenWSN是一个实现相关IEEE和IETF标准的工业无线传感器网络。仿真结果表明，与原始误差模型相比，所提出的误差模型将估计传输可靠性的准确性提高了多达12%。此外，随着突发损失的增加，估计准确性也有所提高。
