Time-Sensitive Networking for Real-Time Wired/Wireless Communication in the Industrial Automation Field

With the evolution of convergent networks and of the pervasive adoption of mobile and wireless technologies, there is a growing need for scalable and flexible communication infrastructures capable of ensuring deterministic performance. Specifically, these infrastructures must ensure sub-millisecond time synchronization, lower latency, lower jitter, and higher adaptability to variable operating conditions than current networks. These requirements become even more stringent in the context of the Industrial Internet of Things (IIoT), where distributed sensors, autonomous robots, mobile platforms, and real-time controllers must cooperate in a perfectly coordinated manner. Therefore, ensuring safety, stability, and operational continuity in automated systems requires deterministic and reliable communication mechanisms that provide predictable latency, precise time synchronization, and seamless mobility support. However, existing time synchronization protocols and traffic scheduling mechanisms are unable to meet the stringent temporal requirements of modern IIoT and robotic applications. This thesis addresses these limitations by extending Ethernet Time-Sensitive Networking (TSN) principles to the wireless domain and by exploiting the new capabilities introduced by Wi-Fi 7 with Multi-Link Operation (MLO). The first part of the thesis experimentally evaluates the application of two significant protocols of the TSN family, namely IEEE 802.1AS and IEEE 802.1Qbv, to wireless networks, using both real and simulated robotic setups. Key challenges are identified, i.e., bimodal latency behavior and suboptimal scheduling under mobility. Then, a novel measurement methodology is proposed to accurately quantify latency contributions across protocol stacks and wireless interfaces, enabling precise characterization of synchronization performance at each sub-system component. Subsequently, the impact and effects of synchronization errors on distributed control systems are analyzed and a mitigation strategy is proposed and validated through simulations. Building on these insights, a mathematical framework is developed to optimize time-aware scheduling in mixed Time-Sensitive (TS)/Best-Effort (BE) traffic scenarios while accounting for clock offsets. The second part of the thesis focuses on the latest generation of Wi-Fi, i.e., the IEEE 802.11be standard, also known as Wi-Fi 7, and its MLO functionality, being a potential complement or alternative to the traffic scheduling mechanism defined by the IEEE 802.1Qbv standard. Different scenarios have been tested to compare the performance of the two approaches, thus highlighting their strengths and limitations. Furthermore, an Artificial-Intelligence (AI)-based algorithm is proposed to optimize wireless link selection and ensure seamless roaming in dynamic industrial environments, thus addressing the mobility challenges typical of wireless networks currently used in the IIoT.