Abstract: Aiming at the problem that LoRa network for static parameter settings of the end devices leads to less flexibility and low communication efficiency, a dynamic parameter selection method based on orthogonal genetic algorithm to improve the scalability of LoRa networks is proposed. The influence of different LoRa transmission parameters on communication performance is analyzed, then the channel conflict access and link budget models are established respectively. A dynamic resource selection algorithm based on multi-objective orthogonal genetic algorithm is introduced to solve the model, and ultimately the parameters with minimized collision probability are obtained. By simulating the single gateway LoRa network of nearly over 10 000 devices, the experimental results show that the dynamic parameter selection algorithm proposed in this paper can improves the Packet Delivery Rate（PDR） of the network 30%, which significantly improve the adaptability and scalability of the network in different application scenarios.

Key words: Internet of Things, LoRa network, low power wide-area network, orthogonal genetic algorithm, transmission parameter combination

摘要： 针对当前大部分LoRa网络因终端采用静态参数配置而导致的网络灵活性和通信效率不高等问题，提出了一种基于正交遗传算法改善LoRa网络性能的动态参数选择方法。分析了不同参数配置对网络通信性能的影响，针对LoRa网络建立信道冲突和链路预算模型，通过引入基于多目标遗传算法的动态参数选择方法求解该模型，最终获得具有最小冲突概率的参数集。通过对超过10 000台设备的单网关LoRa网络的运行结果表明，所提出的动态参数选择算法可将网络的分组交付率（Packet Delivery Rate，PDR）提高30%，显著提高了网络在不同应用场景下的适应性和扩展能力。

关键词: 物联网, LoRa网络, 低功耗广域网, 正交遗传算法, 参数组合