The factories of the future will be highly digitalized in order to enable flexible and interconnected manufacturing processes. Especially wireless technologies will be beneficial for industrial automation. However, the high density of metallic objects is challenging for wireless systems due to multipath fading. In order to understand the signal propagation in industrial environments, this paper provides results from a number of channel measurement campaigns funded by the German research initiative “Reliable wireless communication in the industry”. We give an overview of different measurement scenarios covering visible light communication and radio communication below 6 GHz. We analyze large and small scale parameters as well as delay statistics of the wireless channels. Finally, we discuss the importance of the results for the definition of industrial channel models.

In industrial environments an increasing amount of wireless devices are used, which utilize licence-free bands. As a consequence of this mutual interferences of wireless systems might decrease the state of coexistence. Therefore, a central coexistence management system is needed, which allocates conflict-free resources to wireless systems. To ensure a conflict-free resource utilization, it is useful to predict the prospective medium utilisation before resources are allocated. This paper presents a self learning concept, which is based on reinforcement learning. A simulative evaluation of reinforcement learning agents based on neural networks, called deep Q-networks and double deep Q-networks, was realised for exemplary and practically relevant coexistence scenarios. The evaluation of the double deep Q-network showed, that a prediction accuracy of at least 98 % can be reached in all investigated scenarios.

The steadily growing use of license-free frequency bands require reliable coexistence management and therefore proper wireless interference classification (WIC). In this work, we propose a WIC approach based upon a deep convolutional neural network (CNN) which classifies multiple IEEE 802.15.1, IEEE 802.11 b/g and IEEE 802.15.4 interfering signals in the presence of a utilized signal. The generated multi-label dataset contains frequency- and time-limited sensing snapshots with the bandwidth of 10MHz and duration of 12.8 μs, respectively. Each snapshot combines one utilized signal with up to multiple interfering signals. The approach shows promising results for same-technology interference with a classification accuracy of approximately 100% for narrow-band IEEE 802.15.1 and IEEE 802.15.4 signals. For cross-technology interference, wide-band IEEE 802.11 b/g signals achieve an accuracy above 90 %.

Novel industrial wireless applications require wideband, real-time channel characterization due to complex multipath propagation. Rapid machine motion leads to fast time variance of the channel's reflective behavior, which must be captured for radio channel characterization. Additionally, inhomogeneous radio channels demand highly flexible measurements. Existing approaches for radio channel measurements either lack flexibility or wide-band, real-time performance with fast time variance. In this paper, we propose a correlative channel sounding approach utilizing a software-defined architecture. The approach enables real-time, wide-band measurements with fast time variance immune to active interference. The desired performance is validated with a demanding industrial application example.

The steadily growing use of license-free frequency bands requires reliable coexistence management for deterministic medium utilization. For interference avoidance, proper wireless interference identification (WII) is essential. In this work we propose the first WII approach based upon deep convolutional neural networks (CNNs). The CNN naively learns its features through self-optimization during an extensive data-driven GPU-based training process. We propose a CNN example which is based upon sensing snapshots with a limited duration of 12.8 μs and an acquisition bandwidth of 10 MHz. The CNN distinguishes between 15 classes. They represent packet transmissions of IEEE 802.11 b/g, IEEE 802.15.4 and IEEE 802.15.1 with overlapping frequency channels within the 2.4 GHz ISM band. We show that the CNN outperforms state-of-the-art WII approaches and has a classification accuracy greater than 95 % for signal-to-noise ratio of at least -5 dB.

Wireless technologies in licence-free bands are already used to an extend, where mutual interference decreases the performance. Therefore, an automated coexistence management shall be implemented, which distributes the resources space, time and frequency in an optimal manner. Firstly, we define appropriate quality-of-coexistence (QoC) parameters which are determined by the distributed wireless systems and transmitted to a central coordination point (CCP). Secondly, we introduce an algorithm where the resources space, frequency, and time are considered as discrete variables. Additionally, channel sensing provides important information for the allocation decision by the CCP. The purpose of the concept is a resource allocation, which reduces the likelihood of interferences by considering all resources.

Wireless technologies in licence-free bands are widely used to an extent, where mutual interference decreases the performance. Therefore, a coexistence management shall be implemented, which distributes the resources space, time and frequency in an optimal manner. In this paper we purpose a centralized coexistence management concept, which is called central coordination point (CCP). This is based on SNMP (Simple Network Management Protocol) as a management protocol and real-time Ethernet as a control channel. Those technologies are common in industrial wireless communication systems. Furthermore, we suggest a wireless coexistence MIB (Management Information Base), which can easily be implemented in such wireless communication systems. In addition, we designed an algorithm for automatic resource allocation.

Stationary radio channel modeling does not fulfill the harsh conditions of industrial wireless environments within shared license-free spectrum bands. Passive signal distortion caused by the frequency- and time-selective channel gain, and active interfering emissions have to be considered for reliable industrial coexistence validation.

Real-time industrial wireless systems sharing a crowded spectrum band require active coexistence management measures. Identification of wireless interference is a key issue for this purpose.We propose an efficient implementation of a wireless interference identification (WII) approach called neuro-fuzzy signal classifier (NFSC). The implementation in Matlab / SIMULINK is based upon the wideband software defined radio Ettus USRP N210. The implementation is evaluated in six selected heterogeneous and harsh industrial scenarios within the license-free 2.4-GHz-ISM radio band with variously combined standard wireless technologies IEEE 802.11g-based WLAN and Bluetooth. The evaluation of the NFSC was performed with a binary classification test with the statistical measurement metrics sensitivity and specificity.

Automation in industrial production is getting more important, so the need for high data rate, low latency and low power wireless communication is growing. This requires efficient use of bandwidth, in particular for unlicensed bands. In this paper, we present a hardware architecture for analysing the channel utilisation, which is needed in real-time coexistence management systems. Additionally, we compare the solution to a software implementation with respect to performance and real-time capability. The hardware implementation supports real-time channel utilisation sensing, calculating Fast-Fourier-Transform and logarithm. As shown in the results chapter, the hardware solution exceeds software approaches by more than five orders of magnitude in terms of throughput, calculating the Fast-Fourier-Transform and logarithm of a 1024 value wide vector in 17.26 mu s at a clock frequency of 60 MHz. The FFT size can be dynamically reconfigured using control signals driven by software, which enables reuse e.g. in multi carrier systems such as GFDM, when coexistence management is not active.

Wireless control systems for factory automation (FA) applications are subject to coexistence impairments, especially in license-free spectrum bands. Evaluating the coexistence impact requires the knowledge of appropriate characteristic parameters and the usage of a suitable simulation method. In this paper we propose an integral approach for the event-based simulation of wireless coexisting close-loop networked control systems (NCSs) for mutual impact evaluations based upon the integration of an ordinary differential equation resolving library within an event-based wireless network simulation framework. The approach is evaluated within a harsh FA scenario of two identical closed-loop NCS with two coexisting wireless technologies IEEE 802.11 and PNO WSAN-FA.

— Caused by increasing demands for wireless devices in the last decade, license-free bands such as the 2.4 GHz ISM band become more and more crowded. As a consequence, interferences are inevitable. To avoid impairments caused by interferences, there have to be mechanisms to prevent coexisting systems to interfere with each other in license-free bands. Especially in industrial application fields, a reliable and deterministic medium access is required. In this paper, we evaluate an inter-system cooperation approach, which is designed to be aware of interferences, to ensure reliable real-time communication and further to protect investment for existing wireless systems. In a selected industrial scenario, it outperforms a non-cooperative approach especially in presence of interference. Further, it ensures a reliable communication for more than 99% of the real-time data packet transmissions also in presence of interference.

Industrial wireless communication in license-free spectrum bands such as the 2.4-GHz-ISM band suffer from motion and multipath effects, which cause a high time- and frequency-variant channel attenuation. Additionally, mutual interference from heterogeneous wireless technologies limits real-time capabilities of industrial wireless technologies. Therefore, performance validations of industrial wireless technologies within appropriate industrial wireless environments are necessary. In this paper, we present the first raw measurement data set publication of an industrial wireless environment characterization in a data repository for free public access to enable transparent industrial wireless technology validation and to enhance their comparability. We characterize the whole license-free 2.4-GHz-ISM band with a time resolution of 110 μs and a frequency resolution of 1MHz in a coexistence scenario with four antennas obstructed by robot arm movements. Additionally, the frequency and time variance of the measured channel attenuations are analyzed.

Industrial Automation (IA) applications require deterministic communication channels to ensure a reliable operation. As the wireless medium is a shared medium used by many other wireless technologies, a deterministic medium access method (MAM) is necessary. An improvement of the coexistence behavior can be achieved by applying adaptive MAMs, but they cannot meet any real-time demands. A promising approach to meet real-time as well as coexistence demands are cognitive MAMs. We evaluate the performance of three different cognitive MAMs which differ in the probabilistic prediction model: Two methods are based on Markov modelling (MM) and one method is based on an auto-regressive (AR) model. The MAMs are experimentally evaluated in a worst case wireless measurement scenario.