Electronics Guide

Design Trade-offs

LPWAN technologies achieve their remarkable range and battery life through deliberate trade-offs. Data rates are limited, typically measured in bits per second rather than megabits. Message sizes are constrained, often to tens or hundreds of bytes. Communication may be asymmetric, with stronger downlink capabilities than uplink or vice versa.

These trade-offs suit applications where sensors report small amounts of data infrequently. A soil moisture sensor sending readings every hour, a parking space detector reporting occupancy changes, or a water meter transmitting daily consumption all fit the LPWAN model perfectly. Applications requiring streaming data, real-time control, or large payloads need different technologies.

Link Budget Advantages

LPWAN technologies achieve extended range through exceptional link budgets, often exceeding 150 dB and reaching above 160 dB for the most robust modes. This compares to roughly 100 dB for Wi-Fi and 130 to 140 dB for conventional cellular. The additional margin translates directly into longer range, better building penetration, or reduced transmit power.

Link budget gains come from several techniques: narrow bandwidth concentrates receiver energy for improved sensitivity, spreading or repetition codes provide processing gain, slow data rates allow longer integration times, and simplified protocols reduce required signal-to-noise ratios.

Spectrum Options

LPWAN technologies operate in either licensed or unlicensed spectrum. Licensed spectrum technologies (NB-IoT, LTE-M) use cellular bands with guaranteed interference protection but require carrier relationships. Unlicensed spectrum technologies (LoRaWAN, Sigfox) use ISM bands, offering deployment freedom but requiring compliance with duty cycle and power restrictions.

Regional ISM band allocations differ. The 863 to 870 MHz band (commonly cited as 868 MHz) serves Europe with duty cycle restrictions, typically 0.1 or 1 percent depending on the sub-band, though a few narrow sub-bands permit up to 10 percent. North America uses the 902 to 928 MHz band with frequency hopping or listen-before-talk requirements rather than a duty cycle cap. Asia has various regional allocations, such as 920 to 925 MHz. These differences affect technology availability and device design across markets.

Duty cycle limits are a hard design constraint, not a guideline. A 1 percent limit permits only 36 seconds of transmission per hour per sub-band, which bounds how many messages a device may send and how long each may be. Because low data rates stretch airtime, the slowest and longest-range settings consume the duty cycle budget fastest, coupling range directly to message frequency. Licensed-spectrum technologies avoid this constraint but substitute carrier dependence and subscription cost.

Network Architectures

LPWAN networks typically use a star topology, with devices communicating directly to gateways or base stations. This differs from mesh networks, where devices relay for each other. Star topology simplifies device design and reduces power consumption by eliminating relay functionality.

Gateways connect the wireless network to backhaul infrastructure, typically using cellular, Ethernet, or Wi-Fi. Network servers manage device authentication, message routing, and integration with application platforms. Cloud-based network servers provide scalability and management simplicity.

LoRa Physical Layer

LoRa (Long Range) is a proprietary spread spectrum modulation technique developed by Semtech. Using chirp spread spectrum (CSS), LoRa encodes information in frequency-varying chirps that sweep across the channel bandwidth. This approach provides exceptional interference immunity and sensitivity.

Spreading factors from SF7 to SF12 trade data rate for range. SF7 provides the highest data rate (approximately 5.5 kbps at 125 kHz bandwidth) but shortest range. SF12 offers the longest range but lowest data rate (approximately 250 bps). Newer transceivers add SF5 and SF6 for still higher rates at short range. Across the full set of spreading factors and bandwidths, LoRa spans roughly 0.3 to 27 kbps. Devices can adapt spreading factor based on link quality, optimizing throughput when close to gateways and extending range when needed.

Each increment of spreading factor roughly doubles the symbol duration, which adds about 2.5 dB of link budget but also doubles airtime. A payload that takes tens of milliseconds at SF7 can occupy more than a second at SF12. Airtime therefore drives both battery drain and duty cycle consumption, and it is the single most important number in LoRaWAN capacity planning.

Bandwidth options of 125 kHz, 250 kHz, and 500 kHz affect data rate and noise performance. Narrower bandwidth improves sensitivity but reduces throughput. Coding rates from 4/5 to 4/8 add redundancy for error correction, further trading throughput for reliability.

LoRa's CSS modulation is remarkably robust against narrowband interference and multipath fading. The spreading gain enables reception of signals well below the noise floor, making LoRa suitable for challenging RF environments.

CSS is no longer the only modulation LoRaWAN carries. The LoRa Alliance added Long Range Frequency Hopping Spread Spectrum (LR-FHSS) data rates to the Regional Parameters specification in revision RP2-1.0.2, announced on 10 November 2020, for the EU868, US915, and AU915 regions. The Alliance credits LR-FHSS with high interference resistance and high spectral efficiency, which raise network capacity on an unchanged gateway footprint and make direct links from end devices to satellites practical. Support is optional and backward compatible with earlier revisions of the regional parameters.

LoRaWAN Protocol

LoRaWAN defines the network protocol running over LoRa modulation. Managed by the LoRa Alliance, LoRaWAN specifies device classes, network architecture, security, and regional parameters. The open specification enables multi-vendor interoperability and was approved by the International Telecommunication Union in November 2021 as Recommendation ITU-T Y.4480, Low power protocol for wide area wireless networks, which distinguishes it from the proprietary LoRa physical layer beneath it. Regional parameter documents define channel plans, power limits, and data rates for each regulatory region, so a device certified for one region generally requires reconfiguration for another.

The specification continues to gain capabilities beyond basic messaging: firmware update over the air (FUOTA) delivers new device images through multicast sessions, relay operation extends coverage to devices that cannot reach a gateway directly, and network-based geolocation estimates device position from gateway timing without a satellite navigation receiver. A LoRa Alliance task force is extending LoRaWAN to satellite links, aiming at coverage where no terrestrial gateway is practical.

LoRaWAN networks consist of end devices, gateways, network servers, and application servers. Gateways receive transmissions from devices and forward them to the network server. Since multiple gateways may receive the same transmission, the network server handles deduplication and selects the best gateway for downlink responses.

Adaptive Data Rate (ADR) automatically optimizes device parameters based on network conditions. The network server analyzes received signals and commands devices to adjust spreading factor and transmit power, maximizing throughput and battery life while maintaining reliable links.

Device Classes

LoRaWAN defines three device classes with different trade-offs between latency and power consumption. Class A devices, the lowest power option, transmit when they have data and open brief receive windows after each transmission. Downlink messages must wait for the next uplink, creating latency but minimizing power consumption.

Class B devices add scheduled receive windows between Class A transmissions. Synchronized to network beacons, these windows enable deterministic downlink latency without continuous reception. Class B suits applications requiring periodic commands or status requests.

Class C devices keep receivers active continuously when not transmitting, enabling immediate downlink communication. This mode suits mains-powered devices or applications requiring real-time control. Power consumption is highest in Class C.

Security

LoRaWAN provides end-to-end encryption using AES-128. Two keys protect different data: the Network Session Key encrypts network-related information, while the Application Session Key encrypts application payload. This separation ensures network operators cannot read application data.

Device activation uses either pre-provisioned keys (Activation by Personalization, ABP) or over-the-air key exchange (Over-The-Air Activation, OTAA). OTAA provides better security by deriving session keys from root keys during join, preventing key extraction from devices from compromising network security.

LoRaWAN 1.1 enhanced security with additional keys, improved join procedures, and protections against replay attacks. Organizations deploying LoRaWAN should use 1.1 features where available.

Deployment Considerations

LoRaWAN networks can be deployed as public services, private networks, or hybrid combinations. Public network operators such as The Things Network, Helium, and various commercial providers offer coverage in many regions. Private deployments provide control and security for enterprise applications.

Gateway placement significantly affects coverage. Elevated positions improve range dramatically. A single gateway can cover several kilometers in rural areas or multiple city blocks in urban environments. Indoor coverage requires additional gateways or careful placement.

Capacity planning must consider duty cycle limits and collision probability. The ALOHA-style access means devices transmit without coordination, leading to collisions during heavy traffic. Practical deployments support thousands of devices per gateway when traffic is properly distributed.

Technology Overview

Sigfox uses ultra-narrowband (UNB) modulation to achieve extreme receiver sensitivity and spectrum efficiency. Messages occupy only 100 Hz of bandwidth, enabling receivers to detect very weak signals. This narrow bandwidth provides excellent interference rejection and allows many simultaneous transmissions across the available spectrum.

Uplink messages are limited to 12 bytes of payload, with devices allowed up to 140 messages per day under typical subscription plans. Downlink is limited to 4 messages per day with 8-byte payloads. These constraints enforce truly minimal data exchange, perfect for simple sensor reporting. The uplink data rate is correspondingly low, 100 bits per second in Europe, with wider channels and a higher rate used in North American and other regional configurations.

Twelve bytes is enough for a compact sensor reading, a status word, and a counter, but only with deliberate payload encoding. Sigfox deployments typically pack fields at the bit level rather than sending text or structured formats, and decode them in the cloud. This discipline is a defining characteristic of designing for the platform.

The Sigfox protocol is extremely simple, minimizing device complexity and power consumption. No pairing, joining, or handshaking is required. Devices simply transmit, and the network receives. This simplicity comes at the cost of guaranteed delivery and bidirectional communication flexibility.

Network Model

Sigfox operates as a network operator, similar to cellular carriers. Users subscribe to connectivity services rather than deploying their own infrastructure. This model provides wide-area coverage without requiring customer network management.

Base stations are deployed by Sigfox or regional operators to provide coverage. The network architecture is fully cloud-based, with data delivered to customer applications via callbacks or API access. Device management and data handling occur in the Sigfox cloud platform.

Coverage varies by region, with a strong presence in Europe and uneven deployment elsewhere. Reported network reach spans roughly 75 countries. After financial difficulties, the original Sigfox operator entered receivership in January 2022, and the Singapore-based company UnaBiz acquired the technology and the French operations in April 2022. UnaBiz now stewards the technology and has released device library code publicly to encourage interoperability and reduce vendor lock-in. Applications relying on Sigfox must verify current coverage in their deployment areas and consider roaming options for mobile use cases.

The episode is a useful reminder that LPWAN technology selection carries business risk alongside technical risk. A single-operator network concentrates dependence on that operator's viability, whereas a privately deployed network trades that exposure for the burden of owning infrastructure. Long-lived deployments should weigh both.

Message Redundancy

Sigfox devices typically transmit each message multiple times (often three) on different frequencies and with timing offsets. This redundancy compensates for the lack of acknowledgment and retransmission mechanisms. Multiple base stations receiving the transmission further improve reliability.

The network combines received copies to maximize message recovery. Time and frequency diversity protect against temporary interference and fading. While individual message delivery is not guaranteed, the redundancy scheme achieves high overall success rates.

Applications

Sigfox excels at applications with small, infrequent data needs: asset tracking reporting location periodically, utility meters sending consumption data, environmental sensors reporting conditions, and alarm systems transmitting alerts. The subscription model and wide-area coverage suit deployments across large geographic areas.

The message limits and one-way-dominant communication make Sigfox less suitable for applications requiring frequent updates, large payloads, or reliable bidirectional control. Understanding these constraints guides appropriate application selection.

Cellular LPWAN Evolution

The 3GPP standards body introduced NB-IoT (Narrowband IoT, Cat-NB1) and LTE-M (LTE for Machines, also called Cat-M1) in Release 13 to address IoT requirements within the cellular framework. These technologies leverage existing cellular infrastructure while optimizing for low power and extended coverage.

Using licensed spectrum provides interference-free operation and guaranteed quality of service. Existing cellular operators deploy these technologies alongside traditional cellular services, offering IoT connectivity through established subscription models.

NB-IoT Characteristics

NB-IoT occupies 180 kHz of bandwidth, equivalent to a single LTE resource block of twelve 15 kHz subcarriers, within a 200 kHz channel raster. It is deployable in several modes: standalone (using unused spectrum, such as a refarmed GSM channel), in-band (within an LTE carrier), or guard-band (in the LTE guard bands). This flexibility allows operators to deploy NB-IoT using available spectrum resources without dedicating new bands to it.

Extreme coverage extension comes from repetition and narrow bandwidth. NB-IoT achieves roughly 20 dB of coverage gain compared with standard LTE, enabling deep indoor penetration and basement-level connectivity. Link budgets approaching 164 dB support ranges of tens of kilometers under favorable conditions.

In the original Release 13 version (Cat-NB1), theoretical peak physical-layer rates reach approximately 26 kbps downlink and 62 to 66 kbps uplink with multi-tone operation, falling to roughly 17 to 20 kbps uplink in single-tone mode. Release 14 raised the maximum transport block size from 680 bits downlink and 1,000 bits uplink to 2,536 bits in each direction, which is what defines the Cat-NB2 category; published peaks for it are about 127 kbps downlink and 150 to 159 kbps uplink. All of these figures are theoretical peaks under favorable coverage. A device that reaches its cell only through heavy repetition delivers a small fraction of them. Half-duplex operation and a simplified protocol reduce complexity and power consumption, while Power Saving Mode (PSM) and extended Discontinuous Reception (eDRX) enable years of battery operation.

NB-IoT does not support connected-mode mobility handover, making it unsuitable for fast-moving devices. Applications include static sensors, smart meters, and building automation, where devices remain stationary.

LTE-M Characteristics

LTE-M operates in 1.4 MHz of bandwidth, providing higher throughput than NB-IoT: roughly 1 Mbps in each direction for a full-duplex Cat-M1 device, and appreciably less for the half-duplex variants that most low-cost modules implement. The bandwidth supports voice (VoLTE) and higher data rate applications, while half-duplex operation and power-saving features hold consumption down.

Unlike NB-IoT, LTE-M supports full mobility including handover between cells. This capability enables asset tracking, fleet management, wearables, and other applications where devices move across cell boundaries during operation.

Coverage enhancement mode extends range beyond standard LTE but provides less extreme gain than NB-IoT. The trade-off between throughput, mobility, and coverage positions LTE-M for different applications than NB-IoT.

Power Saving Features

Power Saving Mode (PSM) allows devices to enter deep sleep for extended periods while maintaining network registration. Devices can sleep for hours, days, or even longer, waking only to transmit data. The network stores downlink messages until the device wakes, avoiding the need for frequent receiver activity.

Extended Discontinuous Reception (eDRX) provides a middle ground between continuous reception and PSM. Devices sleep for configurable intervals up to nearly three hours (for NB-IoT) while maintaining periodic paging windows. This enables latency-bound downlink communication with significant power savings.

Combining these features with appropriate application design achieves battery lifetimes exceeding ten years for typical sensor applications. Power consumption during sleep can be as low as single-digit microamps.

Deployment and Roaming

Cellular IoT deployment uses existing cellular infrastructure, with operators enabling NB-IoT and LTE-M on their networks. Coverage depends on operator rollout priorities, with significant variation between regions and operators.

Roaming agreements between operators can enable international device operation, though roaming support varies. Devices intended for multi-country deployment must consider coverage and roaming availability in each target region.

Module and subscription costs for cellular IoT have decreased significantly, though they typically remain higher than unlicensed LPWAN options. The total cost of ownership comparison must include infrastructure costs for private networks versus subscription costs for cellular.

Network lifecycle deserves explicit attention. 3GPP has continued to maintain NB-IoT and LTE-M alongside 5G New Radio, allowing them to coexist in the same spectrum, so they remain the standardized path for massive machine-type communication. Operator practice is less uniform than the standards, however: the retirement of 2G and 3G networks stranded earlier generations of cellular IoT hardware, and individual operators make their own decisions about which IoT radio technologies to carry forward. A device expected to run for ten years should be evaluated against operator commitments in each target market, not against the existence of the standard alone.

Mioty

Mioty uses telegram splitting to achieve high reliability and capacity. Each message is divided into many short sub-packets spread across time and frequency according to a defined hopping pattern. Combined with forward error correction, this lets the receiver reconstruct a complete message from only about half of the sub-packets, providing strong immunity to interference and collisions and supporting very high device densities.

The ETSI standard TS 103 357, first published in 2018 and revised since, defines the telegram-splitting ultra-narrowband physical layer that Mioty uses. Sub-packets occupy carriers of roughly 2 kHz within a 200 kHz band, and the scheme operates at 868 MHz in Europe and near 915 MHz in North America. Because each sub-packet is brief and the pattern is pseudo-random, interference tends to destroy only a few sub-packets rather than an entire message, which yields both high capacity and resilience in congested spectrum.

Two properties distinguish Mioty from its unlicensed peers. Capacity is exceptionally high, with a single base station handling on the order of millions of messages per day. Mobility is also supported, with devices remaining usable at speeds up to roughly 120 km/h, whereas most unlicensed LPWANs assume stationary or slow-moving endpoints. The trade-off is a smaller ecosystem than LoRaWAN, with fewer module vendors and less tooling.

Weightless

The Weightless specification defines multiple protocols for different requirements. Weightless-W uses TV white spaces for long range. Weightless-N provides ultra-low-cost, uplink-only communication. Weightless-P offers bidirectional communication in sub-gigahertz bands.

Weightless adoption has been limited compared with LoRaWAN and Sigfox, though the specifications address valid use cases. Organizations evaluating LPWAN options should weigh Weightless against more established alternatives and assess the maturity of its ecosystem.

RPMA (Ingenu)

Random Phase Multiple Access (RPMA) operates in the 2.4 GHz band, distinguishing it from sub-gigahertz alternatives. The higher frequency provides different propagation characteristics, allowing smaller antennas but generally reducing range and building penetration relative to sub-gigahertz options.

RPMA achieves high capacity through direct-sequence spread spectrum with random time and phase offsets. The technology supports larger payloads and more frequent communication than some alternatives, positioning it for applications with moderate data requirements. Commercial momentum has nonetheless been modest compared with cellular IoT and LoRaWAN.

Adjacent Sub-Gigahertz Options

Several technologies sit near the LPWAN category without matching it exactly, and they occasionally serve the same applications better. Wi-SUN builds mesh networks on IEEE 802.15.4g in sub-gigahertz bands and is widely used for utility metering and street lighting, where mesh relaying suits dense, mains-powered nodes and the star topology of classic LPWAN offers less advantage.

Wi-Fi HaLow, defined by IEEE 802.11ah, extends Wi-Fi into sub-gigahertz spectrum for roughly kilometer-scale links at data rates far above LPWAN levels. It suits cameras and richer sensors that need more than a few bytes per message, at the cost of higher power consumption than a true LPWAN endpoint. Within cellular, 3GPP reduced-capability (RedCap) devices, standardized in Release 17 and extended in Release 18, likewise occupy a middle tier: more capable and more power-hungry than NB-IoT, cheaper and simpler than full 5G. Applications that strain LPWAN payload or rate limits often belong in one of these categories rather than in a forced LPWAN design.

Amazon Sidewalk

Amazon Sidewalk creates shared neighborhood networks using compatible Echo speakers and Ring cameras as gateways. It combines Bluetooth Low Energy for short-range connectivity with two 900 MHz layers: LoRa for long range and frequency-shift keying for simpler devices. Compatible endpoints reach the internet through nearby gateways, including gateways that belong to other households.

By aggregating consumer hardware rather than deploying dedicated infrastructure, Sidewalk achieved broad reach quickly; Amazon reported coverage of more than 90 percent of the United States population in 2023. The network is a United States deployment, so it is not an option for international products. While it differs from operator-run LPWANs, Sidewalk illustrates a crowd-sourced deployment model in which coverage is a byproduct of consumer device sales. Privacy considerations and dependence on the Amazon ecosystem limit its suitability for many industrial IoT deployments, but the approach continues to shape thinking about community-scale networks.

Comparison Criteria

Selecting the right LPWAN technology requires evaluating multiple factors. Coverage availability in deployment regions often eliminates some options immediately. Data rate and message size requirements must match technology capabilities. Latency constraints affect technology choice, with cellular IoT generally providing lower latency than unlicensed alternatives.

Power consumption differences affect battery life and sizing. Cost considerations include device modules, connectivity subscriptions, and infrastructure for private deployments. Ecosystem maturity influences component availability, tools, and support resources.

Figures vary with region, configuration, and release, so the table indicates relative positioning rather than guaranteed performance. Two entries usually decide the choice on their own: whether the application requires mobility across cells, and whether coverage exists where the devices will live.

Use Case Mapping

Smart metering suits any LPWAN technology, though data volume and update frequency guide selection. Utility-scale deployments often choose NB-IoT or Sigfox for wide-area coverage without infrastructure investment. Campus deployments may prefer private LoRaWAN networks.

Asset tracking divides between static and mobile. Stationary asset monitoring works with any technology. Mobile tracking requires LTE-M for continuous connectivity during movement or accepts periodic position reports with stationary-optimized technologies.

Agricultural monitoring benefits from LoRaWAN's range in rural areas where cellular IoT coverage may be limited. Private network deployment provides independence from operator coverage decisions.

Smart city applications often mix technologies based on specific requirements. Parking sensors may use Sigfox or NB-IoT for simplicity. Lighting control may prefer LoRaWAN for local network management. Environmental monitoring chooses based on sensor density and data requirements.

Hybrid Approaches

Some deployments combine multiple technologies. Devices might use LoRaWAN for routine data and cellular fallback for critical alerts. Gateways aggregating local sensor data over LoRaWAN might backhaul via cellular IoT. Understanding technology complementarities enables optimized hybrid architectures.

Device Design

LPWAN device design emphasizes power efficiency. Careful component selection, efficient power supplies, and aggressive power management maximize battery life. Sleep current often dominates energy consumption for infrequently transmitting devices, making low-power sleep modes essential.

Antenna design significantly affects range and reliability. Sub-gigahertz antennas are larger than 2.4 GHz alternatives, which affects enclosure size. Antenna efficiency and placement relative to ground planes, batteries, and enclosures require careful attention.

Module selection involves trade-offs between integration level, certification status, cost, and power consumption. Pre-certified modules simplify regulatory approval. System-on-chip solutions offer lowest power but require more design effort.

Network Planning

Private network deployment requires coverage planning. Propagation modeling estimates coverage from gateway locations, though site surveys provide more accurate results. Indoor coverage typically requires additional gateways or external antenna placement.

Capacity planning must consider device population, message rates, and duty cycle constraints. LPWAN networks can support thousands of devices per gateway under typical conditions, but specific deployments require analysis of actual traffic patterns.

Redundancy and reliability requirements influence architecture decisions. Multiple gateways provide path diversity for critical applications. Backhaul redundancy ensures connectivity despite individual link failures.

Security Considerations

LPWAN security varies by technology. LoRaWAN provides end-to-end encryption with network and application key separation. Cellular IoT inherits cellular security mechanisms. Sigfox uses simpler security appropriate for its limited data model.

Device provisioning must protect credentials from extraction. Secure element integration provides tamper-resistant key storage. Over-the-air activation reduces exposure of root keys compared to pre-provisioned approaches.

Application-layer security may supplement link-layer protection for sensitive data. Consider the complete security architecture from device through network to application when designing systems.

Testing and Certification

LPWAN devices require regulatory certification for radio operation. Regional requirements differ, with separate certifications needed for each market. Pre-certified modules simplify this process but may still require host product certification.

Network compatibility testing verifies device operation with deployed infrastructure. Public network operators may provide certification programs. Private network testing validates coverage and performance in actual deployment environments.

Battery life validation through measurement and modeling ensures devices meet operational lifetime requirements. Accelerated testing techniques help characterize long-term behavior within practical development timelines.