Load- and Interference-Aware CoATI-Based ADR Schemes for LoRaWAN IoT Networks

Abstract

LoRaWAN technology has become foundational for massive IoT deployments requiring long-range, low-energy communication. Adaptive Data Rate (ADR) algorithms play a critical role in optimizing network capacity, energy efficiency, and reliability. However, existing ADR studies lack comprehensive evaluations across dense-to-large scale scenarios and do not fully address gateway load and interference. This paper presents a robust LoRaWAN ADR schemes, supporting Industrial, Suburban, and Rural–Agricultural scenarios with configurable network densities and interference levels. We benchmark three algorithms: (1) Standard-ADR, the LoRaWAN baseline; (2) CoATI-ADR, a metaheuristic-based, load-aware ADR algorithm that considers link quality and SINR metrics; and (3) Binary-CoATI-ADR, a binary population-based variant of CoATI-ADR that is also loadand interference-aware, but treats interference as an explicit, separately weighted fitness objective while jointly optimizing spreading factor, transmit power, and gateway assignment. Our results show that the proposed CoATI-based algorithms outperform Standard-ADR in single-channel LoRaWAN networks, achieving reductions in average transmission power of up to79.5% and energy savings of up to 31.7%. They maintain network reliability, with Effective Packet Delivery Ratio remaining within ±0.1% of Standard-ADR, and improve fairness by up to 0.34% in dense and large-scale scenarios.