By the end of this section, you will be able to:
Before diving into this chapter, you should be familiar with:
LPWAN technologies are designed for IoT devices that need to send small amounts of data over very long distances while running on batteries for years. Think of them as “super efficient messengers” that can travel far but carry only small notes.
| Term | Simple Explanation |
|---|---|
| LPWAN | Low Power Wide Area Network - sends small data over long distances with tiny batteries |
| LoRaWAN | Long Range WAN - can send data 10+ km on battery power for years |
| Sigfox | Ultra-simple network - very limited messages but works globally |
| NB-IoT | Narrowband IoT - uses cell phone networks for IoT devices |
| 802.15.4 | Standard for low-power wireless - basis for Zigbee and Thread |
“I need to send data from a farm field that is 10 kilometers from the nearest building,” said Sammy the Sensor. “Wi-Fi cannot reach that far!” Max the Microcontroller nodded. “That is where LPWAN technologies come in. LoRaWAN, Sigfox, and NB-IoT are all designed to send small messages over very long distances.”
“LoRaWAN is like a whisper that travels incredibly far,” explained Lila the LED. “It only sends a few hundred bytes at a time, but it can reach 10-15 kilometers in open areas. And Sammy’s battery can last 5 to 10 years because the radio only wakes up for a few seconds each time it transmits.”
“For shorter range, there is IEEE 802.15.4,” said Max. “Zigbee and Thread are built on it. They only reach about 100 meters, but they form mesh networks where devices relay messages for each other, extending the overall coverage.”
Bella the Battery was excited about cellular IoT. “NB-IoT and LTE-M use existing cell phone towers, so you do not need to build your own network infrastructure. NB-IoT is great for simple sensors, while LTE-M supports voice and mobility. And 5G is bringing three new categories: massive IoT for millions of devices, ultra-reliable low-latency for factory robots, and enhanced broadband for cameras!”
Since the traditional frame format of MAC layer protocols was not suitable for IoT low power and multi-hop communications, IEEE 802.15.4 was created with a more efficient frame format that has become the most used IoT MAC layer standard.
| Protocol | Focus | Key Feature |
|---|---|---|
| Zigbee | Home automation | Application profiles, mesh networking |
| 6LoWPAN | IPv6 integration | IPv6 over low-power networks |
| Thread | Smart home | IP-based, Matter compatible (Google/Apple) |
| WirelessHART | Industrial | TDMA scheduling, 99.999% reliability |
| ISA100.11a | Industrial | IPv6 native, protocol tunneling |
Key Features:
This variant shows how to select between Zigbee, Thread, and WirelessHART based on application requirements.
Scenario: A manufacturing plant needs to deploy 150 vibration sensors across a 400m x 300m factory floor with metal machinery and concrete walls causing significant RF interference. Sensors transmit 200-byte readings every 5 seconds.
Given:
Analysis:
Result: IEEE 802.15.4 with O-QPSK modulation at 2.4 GHz is optimal. DSSS provides interference immunity, mesh capability extends coverage, and 250 kbps far exceeds the 320 bps requirement while allowing protocol overhead.
Key Insight: In high-interference industrial environments, choose modulation with spread spectrum techniques (DSSS, FHSS) rather than higher-order modulation (64-QAM) which requires cleaner channels. Robustness beats raw throughput for sensor networks.
LPWAN (Low Power Wide Area Network) is a protocol category for resource-constrained devices and networks over long ranges. LPWAN technologies trade bandwidth for range and power efficiency.
LoRaWAN:
Sigfox:
NB-IoT (Narrowband IoT): - Best for: Smart metering, urban deployments, mobile assets - Advantages: Licensed spectrum (reliable), good indoor penetration, mobility support - Disadvantages: Requires cellular infrastructure, subscription costs
LTE-M (LTE Cat-M1): - Best for: Higher data needs, voice capability, mobility - Advantages: Higher throughput, roaming, firmware updates over-the-air - Disadvantages: Higher power consumption than NB-IoT, subscription costs
This variant shows LPWAN technologies through a power consumption lens - useful for understanding why battery life varies dramatically between protocols.
Option A (Licensed - NB-IoT/LTE-M):
Option B (Unlicensed - LoRa/Sigfox):
That 157 dB link budget isn’t marketing fluff—it’s the difference between 10 km rural range and 100 m urban range.
Link budget calculation: \(\text{Margin} = P_{TX} + G_{TX} + G_{RX} - L_{path} - L_{fade} - P_{RX\_sensitivity}\)
For LoRaWAN at 868 MHz, 15 km rural free-space path loss (d in km, f in MHz): \[\text{Path Loss} = 20\log_{10}(d_\text{km}) + 20\log_{10}(f_\text{MHz}) + 32.45 = 20\log_{10}(15) + 20\log_{10}(868) + 32.45 \approx 115 \text{ dB}\]
With 14 dBm TX power, 0 dBi antennas, -137 dBm RX sensitivity (SF12/BW125), and 10 dB fade margin: \[\text{Margin} = 14 + 0 + 0 - 115 - 10 - (-137) = +26 \text{ dB}\]
This is the free-space (best-case) margin. In practice, building penetration, terrain, and vegetation add 25–35 dB of clutter loss. With 30 dB clutter loss added, the effective path loss reaches ~145 dB, reducing margin to −4 dB — confirming that 15 km in dense environments is at the boundary. At 10 km, clutter-adjusted path loss drops to ~141 dB (4 dB improvement), explaining why LoRaWAN is reliably specified at 2–15 km.
Decision Factors: Choose licensed cellular for mission-critical applications (medical devices, utility shutoff valves, security systems) where interference could cause safety/financial harm. Choose unlicensed for cost-sensitive mass deployments (agriculture, environmental monitoring) where occasional packet loss is acceptable and 10-year TCO matters more than guaranteed SLA.
Core Concept: A wireless link works when Received Power exceeds Receiver Sensitivity - calculated as TX Power + Antenna Gains - Path Loss - Fade Margin, where every 6 dB of path loss doubles the required distance or halves the signal strength.
Why It Matters: Before deploying any IoT sensor, you must verify the link will actually work at your target distance. A link budget tells you whether your 10 km rural sensor will reach the gateway, or if you need a higher-gain antenna, more TX power, or a closer gateway.
Key Takeaway: Design for at least 10 dB link margin above receiver sensitivity (-110 to -130 dBm for LPWAN, -70 to -90 dBm for Wi-Fi) to ensure 99%+ reliability through weather, interference, and environmental changes.
Cellular networks are suitable for long-distance communications in IoT applications. Five generations of mobile communication have been developed over the past 30+ years.
GSM is a 2G digital cellular network standard first deployed in 1991, developed under ETSI. It replaced analog 1G systems with digital voice and basic data.
IoT Use: Basic telemetry, SMS-based alerts, simple M2M communication; widely used until carriers began sunset (AT&T retired 2G in 2017, many global carriers by 2025).
3G networks (UMTS/WCDMA) delivered megabit data speeds but were rarely adopted for IoT due to high module costs and power consumption. HSPA and HSPA+ pushed speeds to 21–42 Mbps, but IoT value came from the lower-power 4G derivatives (NB-IoT, LTE-M) rather than 3G itself.
To improve the speed and capacity of cellular networks, LTE based on 4G was standardized by 3GPP (Release 8, 2008).
Cellular networks are expensive technology due to: - Utilization of licensed Radio Frequency (spectrum auction costs) - Intellectual property protection (patent royalties) - Infrastructure deployment and maintenance - Subscription and data fees per device
This makes cellular less attractive for massive IoT deployments where device costs must be minimal.
The 5th Generation cellular network improves IoT communications significantly and promises to:
5G IoT Capabilities:
Scenario: A smart factory deploys three categories of IoT devices on a private 5G network. The network architect must allocate the three 5G service categories (eMBB, mMTC, URLLC) to each application.
Device Categories:
| Application | Devices | Data Rate | Latency | Reliability |
|---|---|---|---|---|
| Quality inspection cameras | 50 | 100 Mbps each | <50 ms | 99.9% |
| Environmental sensors | 5,000 | 100 bytes every 10 min | <10 s | 95% |
| Robotic arm controllers | 20 | 1 Mbps each | <1 ms | 99.9999% |
Step 1: Map Applications to 5G Service Categories
Quality cameras -> eMBB (Enhanced Mobile Broadband)
Reason: High bandwidth (100 Mbps) is the dominant requirement.
Total bandwidth: 50 cameras x 100 Mbps = 5 Gbps
Slice config: 5 Gbps guaranteed, best-effort latency
Environmental sensors -> mMTC (Massive Machine Type Communications)
Reason: Massive device count (5,000) with tiny payloads.
Total traffic: 5,000 x 100 bytes / 600 sec = 833 bytes/sec = 6.7 kbps
Slice config: Minimal bandwidth, connection density focus
Robotic controllers -> URLLC (Ultra-Reliable Low-Latency)
Reason: Sub-millisecond latency and six-nines reliability mandatory.
Total bandwidth: 20 x 1 Mbps = 20 Mbps
Slice config: 20 Mbps guaranteed, <1 ms latency, 99.9999% reliabilityStep 2: Validate Resource Allocation
Available 5G spectrum: 100 MHz at 3.5 GHz (n78 band)
Theoretical capacity: ~1 Gbps per cell (100 MHz, 256-QAM, 4x4 MIMO)
Factory footprint: 3 cells covering 10,000 m^2
Resource allocation per cell:
URLLC slice: 7 Mbps reserved (low bandwidth, highest scheduling priority)
eMBB slice: 1.7 Gbps reserved (5 Gbps / 3 cells)
mMTC slice: 3 kbps reserved (near-zero bandwidth, connection management focus)
Remaining: ~300 Mbps buffer for burst traffic
URLLC feasibility check:
Radio latency (air interface): 0.5 ms (mini-slot scheduling)
Core network latency: 0.2 ms (MEC edge deployment)
Total: 0.7 ms < 1 ms requirement (passed)Step 3: Cost Comparison vs Alternatives
| Solution | Monthly Cost | Meets All Requirements |
|---|---|---|
| Private 5G (n78) | $15,000 | Yes – all three slices |
| Wi-Fi 6E + Ethernet | $3,000 | No – cannot guarantee <1 ms for robots |
| LoRaWAN + Wi-Fi + TSN | $5,000 | Yes, but 3 separate networks to manage |
Key Insight: Private 5G costs 3-5x more than alternatives but is the only single-network solution that simultaneously supports massive device density (mMTC), high-bandwidth video (eMBB), and deterministic sub-millisecond control (URLLC). The cost premium is justified when operational simplicity of managing one network outweighs the capital savings of three separate technologies.
An LPWAN device transmitting every second violates duty cycle regulations and depletes the battery in days, not years. Fix: design applications to transmit at most every few minutes, aggregating readings locally if needed.
LoRaWAN is optimised for uplink (sensor to cloud). Downlink (cloud to sensor) is severely limited by duty cycle. Fix: design applications to be primarily uplink-driven; use downlink only for rare configuration changes or critical commands.
NB-IoT requires licensed spectrum and a cooperating mobile operator. Coverage varies significantly by country and region. Fix: verify NB-IoT coverage at the deployment location with the target operator before selecting NB-IoT as the connectivity technology.
IEEE 802.15.4 Protocols: | Protocol | Best For | Key Feature | |———-|———-|————-| | Zigbee | Home automation | Application profiles, large ecosystem | | Thread | Smart home | IPv6 native, Matter compatible | | WirelessHART | Industrial | TDMA, deterministic timing | | 6LoWPAN | IP integration | IPv6 over low-power networks |
LPWAN Comparison: | Technology | Range | Data Rate | Battery Life | Cost Model | |————|——-|———–|————–|————| | LoRaWAN | 2-15 km | 250 bps-50 kbps | 10+ years | Private network | | Sigfox | 10-50 km | 100 bps | 10+ years | Subscription | | NB-IoT | 1-10 km | 26 kbps DL / 62 kbps UL | 3-10 years | Cellular subscription | | LTE-M | 1-10 km | ~1 Mbps | 1-5 years | Cellular subscription |
Cellular Evolution for IoT:
Selection Criteria:
| Topic | Chapter | Description |
|---|---|---|
| Network Classification: PAN, LAN, WAN | network-physical-classification.html | Understand how 802.15.4, LPWAN, and cellular map to PAN, LAN, and WAN categories in the OSI model |
| Network Topologies | ../network-topologies/network-topologies-intro.html | Explore mesh, star, tree, and hybrid deployments and how they apply to low-power sensor networks |
| LoRa and LoRaWAN Deep Dive | ../lorawan/lorawan-intro.html | Detailed coverage of LoRa chirp spread spectrum modulation, LoRaWAN architecture, and device classes |
| Cellular IoT Deep Dive | ../cellular-iot/cellular-iot-intro.html | NB-IoT and LTE-M in depth: PSM, eDRX, coverage classes, and deployment planning for cellular IoT |
| Zigbee, Thread, and Matter | ../zigbee-thread/zigbee-intro.html | Detailed look at IEEE 802.15.4-based mesh protocols, the Matter standard, and smart home integration |
| WSN: Wireless Sensor Networks | ../wsn/wsn-intro.html | System-level design of multi-hop sensor networks including routing, energy harvesting, and scalability |