International Conference · 2026
Environment-Aware Modeling and Analysis for LoRa Emergency Network Scheduling (poster)
Abstract
Reliable post-disaster communications require scalable ad-hoc networking under scarce control channels. We propose an environment-aware distributed scheduling method for LoRa emergency networks that autonomously avoids collisions and prioritizes urgent messages. Each node probabilistically estimates others’ positions and priorities using pre-computed 3D propagation and hazard maps to optimize its transmission schedule. In urban ray-traced simulations, the method achieves up to a seven-fold improvement in throughput over existing methods while maintaining low collision rates. The scheme further preserves priority-aware latency differentiation. These results indicate robust and scalable emergency communications without centralized control.
Research Note
In the hours after a disaster the cellular base stations are often down, and what remains useful is ad-hoc radio between devices themselves. LoRa suits that role because it reaches several kilometres at low power, but the same design gives it a low data rate and leaves almost no room for a control channel through which nodes could negotiate. With no central coordinator and many nodes transmitting on their own initiative, collisions destroy messages that would otherwise have arrived.
Scheduling is the mechanism that decides who transmits when. Where a coordinator exists it can simply hand out slots, but after a disaster that option is gone, so each node has to infer the situation around it and choose its own timing.
The proposed method gives each node two pieces of prior knowledge, a precomputed three-dimensional propagation map and a hazard map. From these it estimates probabilistically where the other nodes are and how urgent their traffic is likely to be, and optimises its own transmission schedule accordingly. Because building geometry largely determines how far a signal travels, this geographic knowledge substitutes for the communication that the nodes cannot afford.
In ray-traced urban simulations the method reached up to a sevenfold throughput improvement over existing approaches while keeping collision rates low, and it preserved latency differentiation according to message priority.
Related Projects
