Adaptive Channel Map for Time Slotted Channel Hopping Industrial Wireless Networks ⋆ Max Feldman * Gustavo Cainelli * Gustavo Kunzel * Ivan Muller * Carlos Eduardo Pereira * * Universidade Federal do Rio Grande do Sul (UFRGS), Porto Alegre, Rio Grande do Sul, Brasil (e-mail: max.feldman@ufrgs.br, gustavo.cainelli@gmail.com, gustavo.kunzel@ufrgs.br, ivan.muller@ufrgs.br, cpereira@ece.ufrgs.br). Abstract: The use of wireless networks in industrial environments is a reality today because of its advantages, but with the use of such networks, the problem of coexistence becomes inevitable. This paper presents a system to deal with the problems brought on by the networks coexistence, and thus avoid a reduction in the robustness of the network in which this issue has been accomplished. An adaptive channel mapping system is proposed in industrial wireless networks, where the affected channels are removed from the channel map used. A case study of the adaptive channel mapping system in a Wireless HART network is performed. Keywords: Industrial automation, industrial wireless networks, coexistence, channel map. 1. INTRODUCTION The Industrial Wireless Networks (IWN) reached high application over the years, due to its main characteristics such as reduced infrastructure, ease of installation and maintenance, besides the impetus generated by the indus- try 4.0, see Sha et al. (2017). Among the existing wireless industrial communication protocols, the most commonly used are Wireless HART (WH), ISA100.11a and WIA-PA, see Wang and Jiang (2016). However, wireless industrial communication networks present undesirable characteris- tics, as the latency of communications, energy limitations and physical layer failures, see K¨ unzel et al. (2018). Another problem that arises with the use of wireless net- works is the coexistence. When employing this kind of network in industrial environments applications, coexis- tence usually occurs, because of the presence of different networks that share the same frequency spectrum, such as networks based on the IEEE 802.15.4 standard, like WH, and IEEE 802.11 (WiFi), see Ben Yaala et al. (2016). These networks share the industrial, scientific and medical (ISM) frequency band, see Winter et al. (2014). There are techniques that can jointly minimize the problems gener- ated by the networks coexistence, of which we can mention Time Slotted Channel Hopping (TSCH), Direct Sequence Spread Spectrum (DSSS), Time Division Multiple Access (TDMA) and also blacklisting, see HART Communication Foundation (2009). Commonly IWN natively use TDMA, DSSS and TSCH. TSCH is a channel access method with the objective of guaranteeing the diversity of spectrum use, where on each communication is calculated a new channel ⋆ This study was financed in part by the Coordena¸c˜ao de Aper- fei¸coamento de Pessoal de N´ ıvel Superior - Brasil (CAPES) - Finance Code 001. to be used. The choice of the active channel at any given moment is dependent on the instant of time the network is in, the channel offset of the link to be used and the channel map. Blacklisting techniques are a possible improvement and this work will explore them and present an adaptive channel mapping system. Channel mapping can be divided according to some char- acteristics, among them the most common ones are: static/dynamic mapping and centralized/distributed map- ping. The mapping presented in this work is dynamic, that is, it varies with time, and centralized, where only one channel map is used in all network devices. Some advantages of this dynamic and centralized approach are dynamic interference aware and low complexity, hence low computational cost. In industrial environments, it is possible that there are both permanent interferences, as well as intermittent interference, and in these situations, the use of a static channel map will not be appropriated. For adaptive channel mapping application, it is necessary to observe in which channels interferences are present. This can be done directly, through the channels Energy Detec- tion (ED) when the network is not in use, or indirectly, measuring the transmission failures for each channel indi- vidually, and thus infer that the channels with the highest loss rate are the channels most affected by interference. In this work, the spectrum sensing is done by the ED achieved in a distributed and collaborative way. The se- lection of the worst channel is realized by the k-worst algorithm, and the updating of the channel map in all the network devices is done using the command designed for this function. As presented, this paper shows a full adap- tive channel mapping system, from environment analysis to channel map updating, aiming to change the channel Preprints of the 21st IFAC World Congress (Virtual) Berlin, Germany, July 12-17, 2020 Copyright lies with the authors 8335