Source-linked AI summary
Understanding the limits of LoRaWAN
Ferran Adelantado, Xavier Vilajosana, Pere Tuset-Peiro, Borja Martinez, Joan Melia, Thomas Watteyne
TL;DR
LoRaWAN supports large-scale, simple IoT connectivity, but its limitations are not fully understood and can lead to inappropriate use. The paper therefore analyzes its capabilities and limitations across network behavior and application use cases. It concludes that LoRaWAN’s suitability depends on use-case requirements, including device count, spreading factors, channels, ALOHA access, and duty-cycle constraints.
Problem
LoRaWAN’s rapid adoption has outpaced understanding of its limitations, creating a need to clarify where the technology fits IoT use cases.
Method
The article provides a comprehensive, fair, and independent analysis of LoRaWAN capabilities and limitations across network capacity, scale, and application use cases.
Results
LoRaWAN suitability depends on the combination of end-device count, spreading factors, channels, ALOHA access, and maximum duty-cycle regulation for each use case.
Takeaways & Limitations
LoRaWAN is not a single solution for all connectivity needs and must be carefully dimensioned to meet each use case’s requirements.
Takeaways & Limitations
The analysis focuses on LoRaWAN network scale in relation to data rate and duty-cycle regulations.
Abstract
from arXiv · showhide
The quick proliferation of LPWAN networks, being LoRaWAN one of the most adopted, raised the interest of the industry, network operators and facilitated the development of novel services based on large scale and simple network structures. LoRaWAN brings the desired ubiquitous connectivity to enable most of the outdoor IoT applications and its growth and quick adoption are real proofs of that. Yet the technology has some limitations that need to be understood in order to avoid over-use of the technology. In this article we aim to provide an impartial overview of what are the limitations of such technology, and in a comprehensive manner bring use case examples to show where the limits are.
I. INTRODUCTION
LPWAN deployments support large-scale IoT through low-rate, long-range connectivity and simple star networks, but LoRaWAN’s capabilities and limitations require careful matching to use cases.
- LPWAN combines low data rates with robust modulation to provide multi-kilometer communication ranges.
- Simple star topologies enabled by LPWAN simplify network deployment and maintenance.
- LoRaWAN offers raw data rates up to 27 kbps with LoRa or 50 kbps with FSK, and a gateway can collect data from thousands of kilometer-distant nodes.
- The article provides a comprehensive, fair, and independent analysis of LoRaWAN capabilities and limitations across application use cases.
- The paper covers LPWAN technologies, LoRaWAN details, network capacity and scale, applicable use cases, open challenges, and conclusions.
II. OVERVIEW OF LPWAN AND CELLULAR TECHNOLOGIES FOR IOT
The LPWAN landscape includes LoRaWAN, SigFox, Ingenu, and Weightless standards, which trade off data rate, range, power consumption, openness, and network control.
- LPWAN alternatives include Ingenu, Weightless W, N, and P, alongside LoRaWAN and SigFox.
- Ingenu uses RPMA at 2.4 GHz, reaching up to 624 kbps uplink and 156 kbps downlink, but with higher energy consumption and shorter 5–6 km range.
- Weightless-W supports bidirectional communication in TV white spaces, with data rates from 1 kbps to 1 Mbps and battery lifetimes around 3–5 years.
- SigFox uses proprietary ultra-narrowband access and supports about 100 bps uplink, 12-Byte maximum payloads, and at most 14 packets per device per day.
- SigFox’s restrictions and operator-owned business model have shifted interest toward LoRaWAN, considered more flexible and open.
B. Cellular solutions for IoT
Cellular IoT standards offer varied data rates and device complexity, while LoRaWAN provides low-power, long-range connectivity through a star-of-stars architecture and differentiated device classes.
- 3GPP’s IoT approach comprises eMTC, NB-IoT, and EC-GSM-IoT standardized for low-cost, low-complexity machine-type communications.
- eMTC reaches up to 1 Mbps in both directions, NB-IoT up to 250 kbps, and EC-GSM-IoT provides 70–240 kbps data rates.
- LoRaWAN uses a star-of-stars topology in which gateways relay single-hop end-device messages to a central network server.
- Class A devices minimize power consumption but permit downlink only after a successful uplink, whereas Class B adds beacon-synchronized receive windows for downlink traffic.
- LoRaWAN data rates range from 0.3 kbps to 27 kbps depending on the spreading factor.
- A 1% duty-cycle applies to end-devices in EU 868, constraining channel occupancy in unlicensed bands.
IV. CAPACITY AND NETWORK SIZE LIMITATIONS
The paper studies LoRaWAN network scale in relation to data rate and duty-cycle regulations, identifying these as capacity constraints.
- The section analyzes LoRaWAN network scale with respect to data rate, duty-cycle regulations, and related factors.
A. Network size limited by duty-cycle
LoRaWAN capacity is constrained by duty-cycle limits, increasing Time on Air at higher spreading factors, and collisions as offered traffic and network size grow.
- Duty-cycle constraint: 1% duty-cycle in EU 868 permits only 36 sec/hour of transmission per end-device in each sub-band.The required silent off-period is Ts = Ta(1/d − 1).
- Duty-cycle constraint: Higher spreading factors extend communication range but increase Time on Air and the subsequent off-period.Large spreading factors are also used more often than small ones in the described scenario.
- ALOHA capacity: LoRaWAN capacity is approximated as independent ALOHA networks per channel and spreading factor, with packet rates bounded by nd/Ta_i.A collision occurs when simultaneous transmissions select the same spreading factor and channel.
- ALOHA capacity: The number of successfully received packets decreases when end-devices transmit at their maximum packet rate because collisions increase.The cited scenario uses n = 3 channels.
- Scale effects: For 250, 500, 1000, and 5000 end-devices using n = 3 channels, low transmission rates are collision-limited while high rates are duty-cycle-limited.For smaller deployments, the duty-cycle constraint limits maximum throughput.
- Scale effects: Maximum throughput per end-device falls as the number of end-devices grows.
B. Reliability and Densification drain Network Capacity
Reliability mechanisms and densification can consume LoRaWAN capacity: downlink acknowledgments impose additional duty-cycle off-periods, while shared ALOHA access complicates coordination at scale.
- Reliability overhead: Acknowledgments reduce network capacity through both downlink transmissions and the required gateway off-period under duty-cycle regulation.The paper therefore recommends minimizing acknowledged frames in network and application design.
- Reliability overhead: The capacity drain from acknowledged frames calls into question ultra-reliable services over large-scale LoRaWAN networks.
- Densification and coordination: Shared infrastructure combines applications with different reliability, latency, and transmission-pattern requirements over ALOHA-based access.Coordinating these diverse requirements is identified as a main future challenge.
- Densification and coordination: Unplanned gateway deployment and alternative LPWAN deployments could decrease capacity through collisions and the use of larger spreading factors under higher interference.
V. USE CASES
LoRaWAN suits dispersed applications with sparse, delay-tolerant traffic when gateways cover the end-devices, but its contention and airtime limits constrain real-time operation.
- Suitable use cases: Agriculture, leak detection, and environment control generate few periodic or aperiodic messages, tolerate delay, and require long communication range.
- Suitable use cases: LoRaWAN meets these applications’ requirements when gateway deployment provides coverage for all end-devices.
- Real-time limitations: Industrial automation, critical infrastructure monitoring, and actuation require low latency and bounded jitter for real-time operation.
- Real-time limitations: Even 10-Byte packets have around 40 ms of Time on Air at SF=7, while industrial control loops may require 1 ms to 100 ms response times.
- Real-time limitations: ALOHA contention prevents deterministic operation, although small LoRaWAN networks can support applications sampling data every second.The proposed design considerations are minimizing spreading factor and carefully sizing the number of channels, but latency remains nondeterministic.
B. Metering
Metering spans requirements from time-sensitive operation to best-effort monitoring, so each application's requirements must be matched against LoRaWAN's suitability.
- Metering scenarios range from time-sensitive operation to best-effort monitoring.
- Existing application layers can preserve much of the firmware and ecosystem during migration to LPWAN.Examples include Wireless M-Bus for water or gas metering, KNX for building automation, and ModBus for industrial automation.
C. Smart City Applications
LoRaWAN fits smart-city applications with delay-tolerant, periodic or event-driven data, but not applications requiring stringent delay performance or high bit rates.
- Smart parking, waste collection, and lighting succeed with periodic or event-triggered messages and delay tolerance.These applications can involve few tens of messages per day, although simultaneous triggers can create message avalanches.
- LoRaWAN handles wide coverage and many users for smart lighting at the expense of more collisions, latency, and jitter.
- LoRaWAN is unsuitable for transportation applications with stringent delay constraints, while fleet control and management can be supported.
- IP surveillance video requires 130 kbps to 4 Mbps, exceeding LoRaWAN's 0.3–50 kbps per-channel range.
VI. OPEN RESEARCH CHALLENGES
The authors identify capacity, coordination, access, localization, spectrum, topology, and densification challenges needed to extend LoRaWAN beyond current limitations.
- Capacity and coexistence: Duty-cycle effects jeopardize the actual capacity of large-scale LoRaWAN deployments.A 30-second-per-day Time on Air policy can support more than 1000 end-devices per gateway but lacks flexibility for varying network and application conditions.
- Capacity and coexistence: Capacity optimization and traffic slicing for service-specific requirements remain open research issues.
- Access mechanisms: Adaptive channel hopping should address latency, jitter, reliability, and per-channel noise, while reserving channels for critical retransmissions.Feedback mechanisms between gateways and end-devices are identified as important in the uplink-favored system.
- Access mechanisms: Hybrid or complete TDMA could support deterministic traffic while respecting duty-cycle constraints and fairness with co-existing networks.
- New capabilities: TDOA-based triangulation is being developed for end-device geolocation, benefiting from large spreading factors and dense gateway deployments.
- New capabilities: Cognitive radio, multi-hop strategies, and network densification are proposed to address spectrum, power, capacity, and coexistence constraints.The multi-hop direction considers reducing transmitted power and spreading factors, while densification raises co-existence challenges in urban areas.
- Capacity and coexistence: Gateway coordination and reconfiguration mechanisms are needed to limit interference and collisions in co-existing deployments.
VII. CONCLUSIONS
LoRaWAN's suitability depends on carefully dimensioning gateways, end-devices, spreading factors, and channels for each use case; current technology cannot guarantee deterministic monitoring or real-time operation.
- Use-case fit depends on the number of end-devices, selected spreading factors, and number of channels under ALOHA access and duty-cycle regulation.A gateway covering tens of kilometers and serving up to thousands of end-devices must be dimensioned against application requirements.
- Deterministic monitoring and real-time operation cannot be guaranteed with the current LoRaWAN state of the art.