Recently, LoRaWAN has been considered a promising technology for large-scale IoT applications owing to its ability to achieve low power and long range communications. However, LoRaWAN is limited using Aloha random access scheme. When in dense scenarios, such scheme leads to a high number of collisions, thus severely impacts the reliability and scalability of LoRaWAN. In this paper, we investigate the impact of scalability and densification of nodes and gateways on the system reliability taking into account the capture effect. We propose an optimization problem to derive the node distribution at different spreading factors (SF) in LoRaWAN networks with multiple gateways. We then introduce an adaptive algorithm that enables to easily implement SF optimization by adjusting the signal-to-noise ratio thresholds. Moreover, the performance of the proposed algorithm is compared with the performance of legacy LoRaWAN and relevant algorithms from the state-of-the-art. Simulation results show that the proposed algorithm significantly outperforms the state-of-the-art algorithms, and improves the throughput and packet delivery ratio of the network.

最近，由于LoRaWAN具有实现低功耗和远程通信的能力，因此被认为是用于大规模物联网应用的有前途的技术。但是，使用Aloha随机访问方案限制了LoRaWAN。在密集场景中，这种方案会导致大量冲突，从而严重影响LoRaWAN的可靠性和可扩展性。在本文中，我们考虑了捕获效果，研究了节点和网关的可伸缩性和致密化对系统可靠性的影响。我们提出了一个优化问题，以导出具有多个网关的LoRaWAN网络中不同扩展因子（SF）的节点分布。然后，我们介绍一种自适应算法，该算法能够通过调整信噪比阈值轻松实现SF优化。此外，将该算法的性能与传统LoRaWAN的性能以及最新技术的相关算法进行了比较。仿真结果表明，该算法明显优于现有的算法，提高了网络的吞吐量和分组传输率。