PaperlayerView exact revision on arXiv

Source-linked AI summary

A Survey on Scalable LoRaWAN for Massive IoT: Recent Advances, Potentials, and Challenges

Mohammed Jouhari, Nasir Saeed, Mohamed-Slim Alouini, El Mehdi Amhoud

arXiv:2202.11082v3cs.NIeess.SP

TL;DR

Massive IoT requires connectivity for very large numbers of low-power devices, but LoRaWAN scalability is constrained by interference, collisions, and related capacity limits. This survey examines those challenges and reviews physical- and MAC-layer solutions, concluding that existing approaches use LoRa features, alternative topologies, and interference-management techniques to address massive connectivity.

  • Problem

    LoRaWAN faces interference, concurrent transmission collisions, and scalability limits when deployed for massive IoT connectivity.

  • Method

    The survey reviews LoRaWAN scalability challenges and state-of-the-art solutions across the physical and MAC layers.

  • Results

    Existing solutions primarily exploit LoRa features, while others use relay or routing protocols and signal-processing approaches to address interference and collisions.

  • Takeaways & Limitations

    Scalable LoRaWAN research spans parameter and channel assignment, carrier sensing, network densification, and alternative communication or interference-management schemes.

Abstract

from arXiv · show

Long-range (LoRa) technology is most widely used for enabling low-power wide area networks (WANs) on unlicensed frequency bands. Despite its modest data rates, it provides extensive coverage for low-power devices, making it an ideal communication system for many internet of things (IoT) applications. In general, LoRa is considered as the physical layer, whereas LoRaWAN is the medium access control (MAC) layer of the LoRa stack that adopts a star topology to enable communication between multiple end devices (EDs) and the network gateway. The chirp spread spectrum modulation deals with LoRa signal interference and ensures long-range communication. At the same time, the adaptive data rate mechanism allows EDs to dynamically alter some LoRa features, such as the spreading factor (SF), code rate, and carrier frequency to address the time variance of communication conditions in dense networks. Despite the high LoRa connectivity demand, LoRa signals interference and concurrent transmission collisions are major limitations. Therefore, to enhance LoRaWAN capacity, the LoRa Alliance released many LoRaWAN versions, and the research community has provided numerous solutions to develop scalable LoRaWAN technology. Hence, we thoroughly examine LoRaWAN scalability challenges and state-of-the-art solutions in both the physical and MAC layers. These solutions primarily rely on SF, logical, and frequency channel assignment, whereas others propose new network topologies or implement signal processing schemes to cancel the interference and allow LoRaWAN to connect more EDs efficiently. A summary of the existing solutions in the literature is provided at the end of the paper, describing the advantages and disadvantages of each solution and suggesting possible enhancements as future research directions.

I. INTRODUCTION

Massive IoT networks connect very large numbers of devices that send small amounts of sensing data, creating demanding connectivity and deployment requirements. LPWAN technologies address these requirements, but unlicensed-band restrictions, interference, and collisions constrain network operation.

  • Massive IoT is defined by scale, with connected devices ranging from hundreds to billions and primarily transmitting small sensing-data volumes.
  • LPWAN is preferred for IoT connectivity because it combines long communication range, low energy consumption, and low cost.
  • NB-IoT and LTE-M provide higher data rates, bandwidth, and QoS through licensed-spectrum operation, but require complex protocols, higher energy consumption, and greater costs.
  • Unlicensed-band technologies such as SigFox and LoRa impose duty-cycle restrictions that limit end-device uplink transmission time.
  • A 1% SigFox duty cycle permits 36 seconds of channel access per hour and ultimately limits each end device to 140 messages per day after four protocol-reserved messages.
  • LoRa is identified as suitable for massive IoT because it supports low-cost deployment, remote maintenance, worldwide availability, and private networks without external operators.
  • Multiple IoT devices may simultaneously access and transmit on LoRa networks, creating sporadic congestion and packet collisions that motivate scalability research.

A. Related Surveys

Existing LoRaWAN surveys address architecture, security, routing, PHY behavior, ADR, use cases, and LPWAN comparisons, but they leave massive-IoT scalability insufficiently examined. This survey focuses on scalability and recent solutions for efficient deployment.

  • Scope of Existing Surveys: Prior surveys cover LoRaWAN architecture, attacks, vulnerabilities, routing, and multi-hop challenges.The cited reviews discuss components, security countermeasures, protocol vulnerabilities, routing, and multi-hop effects on PHY layers and energy consumption.
  • Scope of Existing Surveys: Other studies examine LR-FHSS, LPWAN technology comparisons, UAV-based LoRa, ultra-dense networks, ADR, and vehicular communication.These works span physical-layer extensions, mobility, deployment modeling, simulation tools, parameter tuning, and mobile-node adaptation.
  • Scope of Existing Surveys: General LoRaWAN surveys also consider confirmed traffic, application fields, PHY and network-level performance, and comparisons with NB-IoT.Comparative studies report LoRa advantages in battery life, capacity, and cost under unlicensed-spectrum operation.
  • Identified Gap: Existing survey categories commonly compare LPWAN technologies using shared metrics while omitting LoRaWAN spreading-factor and adaptive-data-rate considerations.The literature also lacks investigation of recent scalability challenges associated with massive-IoT deployment.

B. Motivation

The paper is motivated by the absence of a current survey dedicated to LoRaWAN scalability in massive-IoT deployments. It systematically reviews primary studies and organizes scalability solutions across PHY, MAC, applications, and future research.

  • Motivation: No current survey discusses LoRaWAN scalability issues in massive-IoT deployment despite their importance to IoT applications.Existing surveys instead emphasize energy efficiency, architecture, reliability, coverage, mobility, or comparisons among LPWAN technologies.
  • Review Method: The authors conduct a systematic literature review to select, evaluate, and analyze primary studies on scalable LoRaWAN deployment.The process classifies articles, then differentiates selected papers through deep reading and content analysis.
  • Review Method: The selected literature is classified into surveys, background, PHY-layer, MAC-layer, and massive-IoT use-case categories.This classification separates foundational LoRaWAN work from physical-layer, protocol, and application-level scalability studies.
  • Paper Scope: The paper reviews LoRaWAN features and scalability solutions from both PHY and MAC perspectives.Its scope includes LoRa system characteristics, LoRaWAN architecture, end-device classes, and joining procedures.
  • PHY-Layer Scalability: PHY-layer interference solutions are organized into spreading-factor assignment, interference cancellation, and gateway densification.The review addresses simultaneous transmissions using identical LoRa PHY-parameter combinations.
  • MAC-Layer Scalability: MAC-layer collision-avoidance solutions include windowing, channel allocation, alternative topologies, carrier sensing, and slotted MAC protocols.These schemes target collided concurrent LoRa transmissions in dense deployments.
  • Applications and Outlook: The paper surveys massive-IoT use cases including space-to-ground communication, LoRa localization, smart buildings, and environmental monitoring.It also discusses solution drawbacks, open issues, and future research directions for ultra-dense IoT networks.

II. BACKGROUND OF LORAWAN

This section introduces LoRa and LoRaWAN as complementary PHY and MAC components and outlines the physical mechanisms governing coverage, data rate, sensitivity, and capacity. It also presents the protocol structure and transmission trade-offs relevant to dense IoT deployments.

  • Architecture: LoRa provides physical-layer modulation, while LoRaWAN supplies the MAC protocol for low-power wide-area networking.Together, they form a low-power and cost-effective wide-area network.
  • Propagation Models: Propagation models estimate received signal strength and help planners predict end-device coverage and the number of required gateways.The paper identifies propagation modeling as a critical part of efficient deployment planning.
  • Channel Characterization: SNR thresholds vary by spreading factor, changing by 2.5 dB for every increase in SF, while receiver sensitivity depends on bandwidth, noise figure, and thermal noise.The sensitivity calculation fixes NF at 6 for gateway hardware and uses −174 dBm as the thermal noise density.
  • Channel Characterization: CSS supports variable transmission rates through spreading factors and bandwidth, with doubled bandwidth doubling the transmission rate and FEC protecting against interference-induced symbol errors.FEC improves protection but reduces data throughput because it adds redundant bits.
  • Channel Characterization: Higher bandwidth increases data rate, whereas higher spreading factor decreases data rate because it supports longer transmission range; LoRa frames contain a preamble, optional header, payload, and optional CRC.The preamble synchronizes receiver and transmitter and can contain 10 to 65,536 symbols.

3) CSS Modulation:

CSS represents information with chirp pulses whose temporal shifts encode symbols, while the spreading factor determines the number of encoded bits and supports configurable range and rate trade-offs. LoRa PHY parameters jointly affect data rate, sensitivity, distance, and energy consumption.

  • CSS Modulation: LoRa CSS enables long-distance communication while resisting interference, fading, and Doppler effects, whereas FSK offers lower power consumption but shorter range.CSS passes the information-bearing signal through forward error correction before transmission.
  • CSS Modulation: LoRa CSS uses rising and falling chirps to transmit data, with each symbol represented by a chirp and its initial frequency indicating the symbol value.The modulator generates chirps with distinct temporal shifts from a base chirp for different digital inputs.
  • CSS Modulation: Bandwidth values of 125, 250, and 500 kHz trade higher data rate for shorter distance, while higher SF lowers the required SNR and increases energy cost for longer-range transmission.For increasing distance, the paper describes lowering bandwidth and increasing SF.

4) Channel Coding:

LoRa channel coding uses a four-stage PHY processing chain to add error protection, whiten and interleave encoded data, and map it into symbols. LoRaWAN then organizes access and connectivity through gateways, network servers, and end devices.

  • Channel Coding: The LoRa transceiver encodes and decodes data through four stages: Hamming-code FEC, whitening, interleaving, and Gray coding.These stages respectively add redundancy, reduce correlation, transpose and shift coded data, and reduce errors between adjacent bits.
  • Channel Coding: Hamming-code FEC uses code rates 4/5, 4/6, 4/7, and 4/8, where redundant bits improve protection against short interference bursts.The chosen code rate determines the encoder output and the number of redundant bits.
  • LoRaWAN MAC: LoRaWAN is a MAC protocol built over LoRa PHY that uses pure Aloha channel access and requires end devices to follow a duty-cycle constraint.The paper relates Aloha usage to skipping listen-before-talk for energy savings in the presence of hidden nodes.
  • LoRaWAN Architecture: LoRaWAN uses a star-of-stars topology in which end devices connect to gateways, gateways forward packets to a central network server, and the server sends downlink traffic or MAC instructions.End devices do not communicate directly with one another.
  • LoRaWAN Device Classes: Class A, B, and C devices provide progressively more downlink reception opportunities, with Class B and especially Class C requiring more power.Class A opens two receive windows after uplink, Class B adds scheduled windows, and Class C keeps receive windows practically continuous.

2) Standard Adaptive Data Rate (ADR):

Standard ADR adjusts LoRaWAN transmission parameters to reduce end-device energy consumption and maximize throughput. Its performance is limited for mobile devices, crowded networks, and time-varying channels.

  • Standard ADR: ADR adjusts data rate, spreading factor, and transmission power to reduce end-device energy consumption while maximizing throughput.The network server and end device can manage ADR-related transmission parameters.
  • Standard ADR: ADR uses uplink link-budget estimates and SNR thresholds to select transmission settings for end devices.Network-managed ADR relies on previously received uplink packets, whereas blind ADR is performed by mobile end devices.
  • Standard ADR: Mobile end devices require blind ADR because network-managed ADR does not function properly when channel attenuation varies with movement.Mobility also changes topology and link quality between end devices and gateways.
  • Standard ADR: ADR performance decreases in crowded networks and time-varying channels, while LoRa interference and concurrent transmissions constrain system capacity.Dense deployments force many devices to share the communication medium, and standard LoRa modulation supports only a limited number of devices.
  • LoRaWAN Versions: LoRaWAN specifications evolved through multiple releases addressing security, functionality, and protocol performance.The supplied version history covers v1.0, v1.0.2, v1.1, v1.0.3, and v1.0.4.

A. SF Assignment

SF assignment schemes allocate spreading factors using distance, signal conditions, fairness objectives, time sharing, or mobility-aware adaptation. Higher SFs extend range but lower data rate and increase collision probability.

  • SF Assignment: Higher spreading factors support longer-distance transmission but reduce data rate and increase collision probability.SF assignment also interacts with transmission power and code rate through ADR.
  • Optimization Schemes: RSSI- and SIR-based allocation schemes seek improved SF assignment, including heuristic assignments and sequential waterfilling with equalized packet time-on-air.The sequential waterfilling scheme extends earlier heuristic work and targets fairness.
  • Dynamic Schemes: Other approaches share an SF among users for limited durations or proactively reschedule SFs for static and mobile end devices.Game theory is used to manage time-limited same-SF assignments, while mobility-aware schemes reschedule using received signal strength.
  • Channel Conditions: SF, bandwidth, and code rate define regions describing immunity to interference and multipath fading.The regions distinguish immunity to both effects, immunity only to multipath, and susceptibility to both.
  • Legacy Schemes: Legacy schemes assign SFs according to end-device distance from the gateway, using equal-interval, equal-area, or random allocation.The equal-interval and equal-area schemes partition the gateway coverage area differently, while random assignment ignores device location.

B. Interference Cancellation

Interference-cancellation research improves LoRaWAN scalability by decoding overlapping signals, while gateway densification expands coverage and capacity. These approaches retain practical constraints including receiver implementation complexity, gateway duty-cycle limits, and inter-gateway interference.

  • Interference Cancellation: LoRa signals using different spreading factors are not entirely orthogonal, so dense networks remain vulnerable to co- and inter-SF interference.Interference cancellation is proposed to support reliable connectivity under concurrent transmissions.
  • Interference Cancellation: Desynchronized-signal decoding algorithms can improve throughput, while slightly synchronized algorithms decode only the strongest signals.The reported evaluation found the first algorithm improved throughput significantly.
  • Interference Cancellation: Capture effect and successive interference cancellation identify or recover signals from simultaneous transmissions, including weaker packets.The capture effect decodes at least one strongest signal, while SIC helps recover weaker signals.
  • Receiver Design: Receiver designs use synchronization, channel estimation, packet detection, clock tracking, or cancellation to decode overlapping LoRa signals.One complete receiver increased legacy LoRa network capacity by 20, but implementation issues remained a drawback.
  • GW Densification: Gateway densification uses multiple gateways and stochastic-geometry models to analyze coverage and scalability in dense deployments.Multi-cell approaches must avoid interference caused by coordination problems between gateways or other technologies.
  • GW Densification: 90% packet success rate was reported for industrial IoT sensing with multi-gateway LoRaWAN deployment.The analysis included confirmed and unconfirmed traffic and a non-standard channel plan.
  • GW Densification: Increasing gateway density cannot eliminate all scalability issues because gateway duty-cycle restrictions continue to limit downstream traffic.This limitation was identified through NS-3 simulation.

D. Key Insights.

The survey identifies three principal scalability directions: improved SF assignment, interference cancellation, and multi-cell gateway deployment. Each addresses dense LoRaWAN connectivity through different PHY or network-level mechanisms, while introducing corresponding coordination or processing challenges.

  • D. Key Insights: New SF assignment approaches are needed because standard distance-based allocation and ADR have limited capacity under changing density and channel conditions.The survey contrasts distance-based assignment with newer allocation methods designed for scalability.
  • D. Key Insights: Interference cancellation mitigates non-orthogonal SF interference, but requires signal-processing schemes that adapt to concurrent transmissions.Capture effect and SIC can assist recovery of weaker packets, while preambles and chirps support synchronization and signal distinction.
  • D. Key Insights: Multi-cell LoRaWAN replaces the single-gateway star arrangement with gateway densification to support dense deployments and improve scalability.Coverage analysis has used channel models and stochastic-geometry-based probability models.
  • D. Key Insights: Gateway densification must avoid new interference caused by miscoordination among gateways or with other technologies.The survey also discusses logical and frequency channel allocation, transmission scheduling, and MAC-inspired collision-avoidance schemes.

A. Windowing & Channel Allocation

LoRaWAN scalability solutions allocate channels, spreading factors, and resources to reduce collisions, while clustering and multi-hop topologies extend capacity and coverage. These approaches involve assumptions and trade-offs involving orthogonality, energy, throughput, and fairness.

  • Multiple channels can separate devices with different traffic loads, while channel and back-off allocation targets throughput and end-to-end latency.
  • LoRa’s six spreading factors provide bounded virtual channels, limiting simultaneous uplinks and therefore constraining network capacity.
  • Several allocation methods assume perfect spreading-factor orthogonality, while clustering and forwarding must also consider fairness and energy consumption.
  • Resource-block schemes divide channel–spreading-factor combinations so end devices can transmit simultaneously without collision.
  • Logical channels combine bandwidth and spreading factor, producing distinct bit rates that reflect spreading-factor orthogonality.
  • Clustering and multi-hop communication address star-topology limits by improving fairness, reducing redundant transmissions, or extending coverage.Clustering can distribute energy consumption and increase network lifetime, while multi-hop communication can provide high throughput with large coverage.

C. Carrier Sensing

Carrier sensing and listen-before-talk mechanisms are alternatives to LoRaWAN’s collision-prone Aloha access, with reported reductions in collisions and improved channel use. Their deployment remains constrained by sensing energy, hidden nodes, synchronization, and LoRa-specific timing restrictions.

  • Carrier sensing can reduce collisions but increases energy consumption, while slotted MAC designs face difficulty adapting to LoRa PHY characteristics and duty-cycle restrictions.
  • Listen-before-talk reduces simultaneous transmissions on the same channel and spreading factor by checking whether the channel is idle.
  • The binary exponential back-off algorithm achieved minimum channel-activity-detection energy consumption and significantly decreased collision rate.
  • Channel-activity detection was evaluated under laboratory and field conditions, with payload symbols and preambles remaining detectable up to 4 km from a gateway.
  • A scheduled scheme combining distinct spreading factors, frequency channels, time slots, and repeated channel sensing analytically and experimentally achieved superior packet delivery rates.

E. Key Insights.

The survey identifies channel management, topology changes, carrier sensing, and slotted access as routes toward scalable LoRaWAN, while emphasizing goodput, fairness, energy, and realistic orthogonality assumptions. It also reviews space-to-ground LoRa communications, where Doppler and low data rates impose substantial constraints.

  • Throughput alone can overstate LoRaWAN performance because duplicated packets waste bandwidth and energy; goodput is therefore a more appropriate evaluation measure.
  • Future routing protocols should address fairness in throughput and energy consumption, including forwarding-task allocation based on residual energy.
  • Carrier sensing reduces concurrent-transmission collisions but consumes additional end-device energy, motivating hybrid MAC designs.
  • Slotted MAC protocols require adaptation to LoRa PHY properties and duty-cycle restrictions, while synchronization introduces overhead.
  • A. Space-to-Ground Communications: LoRa satellite systems have been studied for large-scale coverage, traffic-load distribution across LEO constellation positions, and CubeSat-based remote connectivity.
  • A. Space-to-Ground Communications: LEO LoRa links experience strong Doppler constraints: simulations reported a maximum shift of about 22 kHz and a maximum Doppler rate of 270 Hz/s.
  • A. Space-to-Ground Communications: An SF of 9 reached a 100% packet error rate under the reported Doppler conditions, while SF 12 caused degradation when the nanosatellite was directly above the ground station.

B. LoRa-based Localization Systems

LoRa-based localization research spans RSSI, range, fingerprinting, filtering, and angle-of-arrival methods, with applications in indoor, outdoor, and smart-building settings. Reported accuracy depends strongly on the method and environment.

  • Recent LoRa localization studies compare systems using performance results summarized in Table X.
  • An indoor RSSI study reported LoRa localization with an average error of 0.62 m, between Wi-Fi’s best and BLE’s worst accuracy.
  • Range-based localization converts gateway RSSI to distance using a path-loss model, whereas fingerprinting compares current RSSI with previously collected values.
  • Bayesian and Kalman filtering address indoor range errors caused by multipath, while particle filtering was applied to LoRa pseudorange fitting.
  • An urban angle-of-arrival system reported average errors of 2° under line-of-sight and 10° under non-line-of-sight conditions.
  • Cell localization can be sufficient when applications need only the room, hall, or place rather than an exact object position.
  • LoRa localization is relevant to smart buildings and healthcare settings where crowded infrastructure and patient movement create tracking demands.

D. Environmental Monitoring

LoRaWAN supports environmental monitoring through wide coverage, low transmission rates, and adaptive communication parameters. However, massive deployments remain constrained by interference, collisions, and limited legacy LoRaWAN features.

  • Environmental monitoring: LoRa is used to monitor harsh environmental conditions because its large coverage and low transmission rates reduce energy consumption.
  • Environmental monitoring: Adaptive spreading factors and data rates allow LoRa communication to respond dynamically to changing environmental conditions and link quality.
  • Environmental monitoring: Environmental monitoring deployments use LoRa end devices and gateways to measure radiation and provide real-time warnings of potential releases.
  • Environmental monitoring: LoRaWAN supports groundwater management with real-time level information, and its low cost and power consumption suit developing regions.
  • Scalability challenges: Massive environmental deployments face interference and scalability challenges because legacy LoRaWAN features and standard ADR provide limited support for dense connectivity.

B. Efficient Interference Cancellation Schemes

The survey reviews interference- and collision-mitigation strategies across LoRaWAN physical and MAC layers. Approaches include signal processing, gateway densification, resource coordination, carrier sensing, and slotted access, each with practical constraints.

  • Physical-layer schemes: Interference cancellation uses FFT, discrete Fourier transforms, capture effects, and preamble detection to decode multiple superposed LoRa signals.
  • Network architecture: Gateway densification, inspired by cellular networks, can increase LoRaWAN capacity in crowded environments but may require architectural changes.
  • Resource management: Resource management aligns end-device transmission windows with gateway reception windows, but density and energy constraints limit its capacity benefits.
  • Network architecture: The conventional star topology creates concurrent gateway collisions that reduce received packet rates and increase latency through retransmissions.
  • MAC-layer schemes: Listen-before-talk improves channel utilization, while carrier-sensing energy costs and long time-on-air limit efficient CSMA deployment.
  • MAC-layer schemes: Slotted MAC protocols can replace poorly scalable pure Aloha-like access, but synchronization, scheduling, duty-cycle restrictions, and decoding overhead remain challenges.
  • Conclusion: The survey identifies interference and concurrent transmission collisions as major scalability problems and reviews PHY and MAC solutions including CSMA and S-MAC.
Loading 2202.11082v3…