Analysis of the Network Attachment Delay of Mobile Devices in the Industrial Internet of Things

Abstract

Industrial networks are typically used to monitor safetyrelated processes where high reliability and an upper bounded latency are crucial. Because of its flexibility, wireless is more and more popular, even for real-time applications. Because radio transmissions are known to be lossy, deterministic protocols have been proposed, to schedule carefully the transmissions to avoid collisions. In parallel, industrial environments now integrate mobile industrial robots to enable the Industry 4.0. Thus, the challenge consists in handling a set of mobile devices inside a static wireless network infrastructure. A mobile robot has to join the network before being able to communicate. Here, we analyze this attachment delay, comprising both the synchronization, and the negotiation of dedicated cells. In particular, since the control frames (EB and 6P) have a strong impact on the convergence, our proposed model carefully integrates the collision probability of these packets. We validate the accuracy of our model, and we analyze the impact of the different EB transmission policies on the discovery delay. Our performance evaluation demonstrates the interest of using efficiently the radio resources for beacons to handle these mobiles devices.

Optimal Initial Synchronization Time in the Minimal 6TiSCH Configuration

“…Specifically, it has been observed that the use of a higher number of channel offsets for EBs decreases the joining time, since it decreases the number of EB collisions, but it decreases the resources available for data transmissions [15]. This consequence has also been validated by other authors [17]. The same studies also observe that negotiating the EB channels, instead of using random ones, leads to increased overhead and, thus, it should be avoided.…”

“…For convenience, we will use the term advertisers to refer to the nodes that send EBs. The standard does not define the rate at which an advertiser sends EBs, but it states that the EB [16,17,23,18,26,15,25,22,19,11,12,13,24,14,20,21]). rate is configured by a higher layer as appropriate taking into account the density of the nodes, the desired time for network formation, and the energy devoted to network formation.…”

“…The problem of the initial synchronization has been studied in the last few years from different perspectives [15], [16], [17]. However, none of these studies takes into account the impact of the scan period on the joining time and on the energy consumption.…”