Time-Constrained Erasure Correction for Data Recovery in UAV-LoRa-WuR Networks

TL;DR

The paper tackles reliable uplink data recovery in UAV-aided LoRa networks that employ Wake-Up Radio, under strict time and energy constraints. It introduces two erasure-correction strategies—random linear fountain coding and message replication—designed to fit within the limited hovering window by distributing redundancy across available uplink slots. An analytical framework is developed to compute the per-message delivery probability, accounting for wake-up dynamics, slot-level collisions, and decoding conditions, and is complemented by numerical results showing when redundancy yields gains. The work provides practical guidance for selecting coding versus replication based on sensor energy budgets, UAV hovering time, and node density, with the framework enabling informed design decisions for reliable data collection in IoT deployments.

Abstract

We described two erasure-correction schemes for data recovery in UAV-LoRa-WuR networks. Our results show that unless the maximum number for redundant frames a sensor can send per data-collection cycle is very small, erasure coding provides noticeable improvements over an uncoded transmissions. Whether to employ coding -- and if so, which type -- should be determined based on the sensors' energy budget (which dictates the maximum redundancy), the UAV's hovering time, and the node density. The analytical framework presented above aids in this decision making.

Figures (3)

• Figure 1: Reliability vs. UAV's hovering time $(\varepsilon \!=\! 4$). Circles show simulated values, lines show analytical results.
• Figure 2: Reliability as a function of UAV hovering time.
• Figure 3: Reliability as a function of node density $n$ ($N_s \!=\! 60$.)

Theorems & Definitions (4)

• Lemma 1
• Lemma 2
• Lemma 3
• Lemma 4