论文标题
大规模LORA网络的分析和优化：吞吐量公平和可扩展性
Analysis and Optimization for Large-Scale LoRa Networks: Throughput Fairness and Scalability
论文作者
论文摘要
Lora网络在广泛领域的低成本和功率约束的用户设备（UES）方面是关键的，而一个关键问题是有效地分配无线资源来支持潜在的大规模UES，同时解决了近乎近乎公平的公平问题，这是由于缺乏易于分析的模型和低频率而挑战的，并且对于较低的模型和低调的指标而言，这是挑战性的。在随机几何形状上，尤其是Poisson雨模型上，我们为总干扰分布，包包成功概率以及带有频率重复使用的多单元格设置的总干扰分布，数据包的成功概率以及系统吞吐量而得出（半）闭合形式公式，并通过计算通道淡出，随机UE分布，部分包装，部分包装，和/或多个计算机，和//或多分配。分析公式仅需要平均渠道统计和空间UE分布，这可以进行可处理的网络性能评估，并孵化我们提出的迭代平衡方法（IB）方法，该方法迅速产生了联合扩散因子（SF）分配，功率控制和占空比的高级政策，以调整Max-Min-min Ue taby Ue ue UE UE UE UE的平均值。数值结果证明了我们提出的优化方案的分析公式和有效性，从而大大减轻了近乎婚姻的公平性问题并减少了空间功耗，同时可以通过适应UE/Gate-ageway eNway centection compations of the Partial offuts。
LoRa networks are pivotally enabling Long Range connectivity to low-cost and power-constrained user equipments (UEs) in a wide area, whereas a critical issue is to effectively allocate wireless resources to support potentially massive UEs while resolving the prominent near-far fairness issue, which is challenging due to the lack of tractable analytical model and the practical requirement for low-complexity and low-overhead design. Leveraging on stochastic geometry, especially the Poisson rain model, we derive (semi-) closed-form formulas for the aggregate interference distribution, packet success probability and hence system throughput in both single-cell and multi-cell setups with frequency reuse, by accounting for channel fading, random UE distribution, partial packet overlapping, and/or multi-gateway packet reception. The analytical formulas require only average channel statistics and spatial UE distribution, which enable tractable network performance evaluation and incubate our proposed Iterative Balancing (IB) method that quickly yields high-level policies of joint spreading factor (SF) allocation, power control, and duty cycle adjustment for gauging the average max-min UE throughput or supported UE density with rate requirements. Numerical results validate the analytical formulas and the effectiveness of our proposed optimization scheme, which greatly alleviates the near-far fairness issue and reduces the spatial power consumption, while significantly improving the cell-edge throughput as well as the spatial (sum) throughput for the majority of UEs, by adapting to the UE/gateway densities.
