A Comparative Evaluation of Multi-Channel Diversity and Random Linear Network Coding for Scalable LoRaWAN under DutyCycle Constraints

LoRaWAN scalability is challenged by ALOHA-based collisions and strict duty-cycle regulations. This study compares two scalability strategies: multi-channel diversity (MC) and random linear network coding (RLNC) under EU868 duty-cycle constraints. Through 1,440 ns-3 simulations across 72 scenarios (50-200 nodes, three traffic levels), we demonstrate that MC consistently improves packet delivery ratio (PDR) by up to 150% and reduces latency by 76%, with no adverse effects on fairness or energy consumption. In contrast, generation-based RLNC fails under EU868 constraints: overhead penalties (5--10 bytes per packet) violate duty-cycle limits, causing PDR to decline by 59-86% and latency to increase more than fivefold. These issues persist even when combining RLNC with multiple channels, demonstrating that per-device duty-cycle limits constitute the fundamental scalability bottleneck. Our analytical model quantifies RLNC failure through multiplicative penalties: overhead-induced rate reduction combined with generation success probability. RLNC may be viable in sparse networks with relaxed duty-cycle regimes, but under dense EU868 deployments with strict 1% duty-cycle limits, RLNC is fundamentally unsuitable. Operators should prioritize enabling all available gateway channels as a first-order optimization, while researchers should treat duty-cycle compliance as a primary constraint when proposing protocol enhancements. All simulation data and outputs are publicly available