Comparative Assessment of Adaptive Data Rate Algorithms for LoRaWAN IoT Applications

Low Power Wide Area Networks (LPWANs) based on the LoRaWAN standard represent a good candidate for Internet of Things (IoT) applications featuring a large number of low-power and low-cost devices distributed over wide areas. The choice of the LoRa radio transmission parameters is crucial for network performance to meet the requirements of IoT applications. The LoRaWAN specification suggests to leverage an Adaptive Data Rate (ADR) algorithm to dynamically assign the transmission parameters to the end-devices, but it does not specify any ADR algorithm. The ADR mechanism manages two transmission parameters, i.e., the transmission power (TP) and the spreading factor (SF). A higher SF reduces the data rate to improve coverage and transmission reliability, at the expenses of longer Time-on-Air (ToA) and higher energy consumption. To maximize the data rate, Semtech recommendations envisage selecting SF and TP based on the most recent link quality measurements relying on the maximum SNR observed over the latest 20 values. Other approaches in the literature, instead of the maximum SNR, use the mean or the median, to provide a more realistic link quality estimation under variable channel conditions. This paper contributes to a better understanding of how to accurately estimate the link quality in LoRaWAN IoT networks providing the results of comparative performance assessments of three ADR algorithms through OMNeT++ simulations. In particular, the paper presents and discusses the results obtained in a scenario with a varying number of end-devices and different configurations of the ADR algorithms.