A Hidden Parameter Study for Traffic-oriented LoRaWAN Deployment

The Long Range Wide Area Network (LoRaWAN) is the leading open protocol reference for Internet of Things operator networks worldwide. Strengthened by the dynamics of its anchoring in a large and very active non-profit community, it offers technological flexibility that can allow it to adapt to the perpetual challenges of the contextual complexity of the IoT environment. The success of the LoRa network is due to the various contributions of improvements to LoRaWAN’s native ADR data rate self-adaptation mechanism. In this paper, after reviewing some improvement proposals, we study how the number of upstream messages used to assess the decision to change node parameters impacts the network performance. We found that minimizing this hidden parameter, called history range, increases the success rate of received packets in the case of a heavy traffic network. Typically, by considering an urban network consisting of a thousand nodes served by five LoRa gateways with a history range varying from 4 to 20, our results show an improvement in the packet delivery ratio metric with a history range of 4. Also, the interference is reduced by up to 42 % in the best case.