Channel assignment in TSCH-based wireless sensor networks using fuzzy logic

Autor: César Benavente-Peces, Diego V. Queiroz, Iguatemi E. Fonseca, Ruan Delgado Gomes, Marcelo S. Alencar
Rok vydání: 2021
Předmět:
Zdroj: Journal of Ambient Intelligence and Humanized Computing. 14:6043-6062
ISSN: 1868-5145
1868-5137
DOI: 10.1007/s12652-020-02741-1
Popis: Recent advances in wireless sensor networks, especially in industry (IWSN), have brought important improvements to deploy a sensor network in such environments. However, there is a high interference level, noise, shadowing, and multipath fading, caused by machines, several metallic objects, which affect the link quality. Some mechanisms have been proposed to deal with these effects, such as frequency hopping and channel blacklisting, or rather, denylisting. However, there are still open issues, such as an appropriate method for managing both mechanisms to avoid similar or adjacent channels in simultaneous transmissions. In addition to denylisting with low quality channels, this paper proposes a greylist with uncertain quality channels, and a whitelist, or rather, allowlist, with good quality channels. A fuzzy logic method to classify them and insert the channels into the appropriate list is used. The mechanism was compared to the channel hopping approach employed on the time slotted channel hopping (TSCH) protocol with and without channel offset management in tree topology networks. The evaluations were made through simulations using a realistic channel model for IWSN, and the results demonstrate that the proposed approach presented better performance in terms of packet delivery rate and determinism. In the studied examples, the network with 53 devices and with the triple channel list approach presented 8% better in terms of packet delivery rate in the application layer than the blind approach (without link quality estimation) of TSCH with channel offset management (no collision), and from 51 to 59% better than approaches without the channel offset management.
Databáze: OpenAIRE