Network Size Estimation in DtS-IoT: A LoRaWAN Approach for Satellite Constellations

Low Earth Orbit (LEO) constellations are pivotal for the global connectivity of Internet of Things (IoT) networks. They facilitate Direct-to-Satellite (DtS) communications for ground nodes and address the absence of terrestrial infrastructure in remote regions. Although LoRa-based solutions provide low-cost connectivity to constrained devices in DtS-IoT networks, scalability remains a significant challenge due to heavy channel congestion associated with large-scale deployments. The LoRa Optimistic Collision Information-based (L-OCI) mechanism has been designed to estimate efficiently the network size of a LoRa-based DtS-IoT deployment. It enables a passing satellite to gauge the potential congestion in the served region by estimating the number of active ground nodes within its coverage area. However, this mechanism was devised to operate assuming a single LEO satellite. This paper introduces the constellation LOCI (CL-OCI), a new methodology for network size estimation designed for LEO satellite constellations. We thoroughly tested CL-OCI using FLoRaSat, a discrete-event satellite constellation simulator with realistic orbital DtS-IoT scenarios. Results indicate a substantial improvement in scalability, doubling the maximum number of nodes CL-OCI can estimate, and a sevenfold reduction in energy wasted due to collided transmissions compared to the baseline single-satellite L-OCI.