Análise de escalabilidade em redes LoRaWAN: um estudo comparativo de estratégias de alocação de spreading factors no NS-3

Abstract

Introduction: The Internet of Things (IoT) drives the massive adoption of LPWAN technologies such as LoRaWAN, highlighted by its robustness and energy efficiency. In high-density scenarios, scalability is a critical challenge: the ALOHA protocol adopted by LoRaWAN increases packet collision probability as device density grows. This challenge is compounded by global regulatory diversity — this work investigates two distinct plans: EU868 (Europe), which imposes a 1% Duty Cycle on gateway transmissions, and AU915 (Brazil), which operates with 64 channels and no such constraint, creating intrinsically different scalability environments. Objective: To compare the performance of two Spreading Factor (SF) allocation strategies — distance-based Static Allocation and the dynamic Adaptive Data Rate (ADR) mechanism — in massive network scenarios under EU868 and AU915. Methodology: A computational simulation study was conducted using NS-3, evaluating five density levels (100 to 5000 nodes) in both regions, with 33 repetitions per configuration for statistical validation. Results and Discussion: In the EU region, ADR suffered severe saturation with 5000 nodes (PDR ≈1.54%), caused by the depletion of the gateway downlink budget — the so-called Bootstrapping Problem. In the BR region (AU915), the same scenario yielded a PDR of 60.41% with ADR and 86.02% with Static Allocation. However, the physical cross-check revealed that real gateway limitations reduce this theoretical ceiling from 86% to 41%. Conclusion: Regionalized planning and awareness of hardware limitations are crucial for the viability of massive LoRaWAN networks.