Towards Dependable IoT via Interface Selection: Predicting Packet Delivery at the End Node in LoRaWAN Networks

first_pagesettings Open AccessArticle Towards Dependable IoT via Interface Selection: Predicting Packet Delivery at the End Node in LoRaWAN Networks by Marc Guerrero 1,*,Cristina Cano 1ORCID,Xavier Vilajosana 1ORCID andPascal Thubert 2ORCID 1 Computer Science, Multimedia and Telecommunications Department, Universitat Oberta de Catalunya, 08018 Barcelona, Spain 2 Cisco Systems France S.à.r.l., 92130 Issy-les-Moulineaux, France * Author to whom correspondence should be addressed. Academic Editor: David Plets Sensors 2021, 21(8), 2707; https://doi.org/10.3390/s21082707 Received: 10 March 2021 / Revised: 28 March 2021 / Accepted: 3 April 2021 / Published: 12 April 2021 (This article belongs to the Special Issue Dependable IoT Networking) Download PDF Browse Figures Review Reports Citation Export Abstract Estimating channel conditions to predict packet delivery can be exploited as a powerful tool to ensure wireless networks dependability. In this article we explore the practical application of this idea from the end-device perspective, using the LoRaWAN protocol stack. We aim to understand if packet delivery can be estimated considering different levels of feedback at the end-device. For that, an extensive data collection campaign is carried out. Through an analysis of the obtained traces, we establish correlations between connectivity metrics at the end node and the fact that a packet is received at the gateway. The study is complemented considering different levels of feedback: (i) No feedback, (ii) enabling acknowledgements frames, and (iii) considering application/control plane data about the channel status at the gateway side. The results show that it is possible to estimate packet delivery in all the evaluated cases.