Industrial Wireless Networks: The Significance of Timeliness in Communication Systems

In the last few years, wireless networks have gained significant importance in the context of industrial communication systems, where their deployment is bringing several noticeable benefits, ranging from replacement of cables to the connection of devices that cannot be reached by traditional wired systems. These features make the adoption of wireless networks for industrial applications very attractive, and they are envisaged to be deployed even more in the future, either as stand-alone systems or arranged in hybrid (wired/wireless) configurations. Unfortunately, wireless communication systems are often characterized by well-known problems, such as fading, multipath propagation, shadowing, and interference, that have the undesired effect of increasing the bit error rate (BER), resulting in the introduction of delays as well as randomness in packet delivery. Moreover, in the context of industrial communication, these aspects may be exacerbated by the specific nature of the environment. Indeed, the rapid movement of machineries along with the possible presence of electromagnetic interference sources, which are typical of manufacturing sites, may introduce considerable fluctuations of the BER values that contribute to further degradation in communication quality. All of these phenomena may have a negative impact on the performance of industrial wireless communication systems, particularly on their timeliness. This is a crucial aspect, since such systems are often required to provide very tight timing performance as dictated by the typical application fields in which they are employed, such as factory automation, process control, and manufacturing systems. Timeliness, in a literal sense, is a general term that indicates the ability to cope with some timing constraints. As such, it has been adopted in the industrial communication scenario. In this context, an actual timeliness assessment can be performed through the evaluation of some performance indicators that are tailored to the specific network employments. In particular, several applications require that some process data are periodically delivered at precise instants (such as, for example, in motion control systems). In this type of application, when communication networks are employed, suitable performance indicators to evaluate their timeliness are represented by both the cycle time values that can be achieved by the networks and the maximum jitter that may affect the execution of cyclic operations. On the other hand, for different kinds of applications, like those concerned with event-driven systems such as remote monitoring, timeliness may be better addressed referring to the maximum delivery time of acyclic data, since this metric allows for estimation of the capability of the network to promptly notify the occurrence of critical, unpredictable events, such as alarms. The analysis we provide is mainly focused on both the physical and the data link layers, since industrial networks are typically based on reduced protocol stacks that, as such, strongly rely on these layers to provide adequate timeliness. Moreover, high-layer protocols adopted by different industrial wireless networks, although not interoperable, frequently rely (at least partially) on common, standardized, low-level protocols [7]. An example in this direction is given by the well-known standards used for process control and building automation, such as wireless highway addressable remote transducer (HART), ISA 100.11a, and ZigBee, which rely on the IEEE 802.15.4 Wireless Personal Area Network (WPAN). Similarly, IPv6 over low-power wireless personal area networks (6LoWPAN), a standard that is beginning to be employed even in the industrial environment, defines an adaptation layer that allows the use of IPv6 communication services on top of IEEE 802.15.4. Further examples of this type are applications in which the authors report the deployment of an IEEE 802.15.4-based sensor network in an oil refinery, and where the authors discuss the adoption of industrial wireless sensor networks (IWSNs) for intrusion detection systems. Finally, it is worth mentioning the popular wireless interface for sensors and actuators (WISA), which implements a master-slave protocol exploiting the physical layer of the IEEE 802.15.1 WPAN (Bluetooth).