11  Wireless Propagation and Design

In 60 Seconds

Signal strength decreases predictably with both distance and frequency according to the free-space path loss (FSPL) formula. Sub-GHz bands (868/915 MHz) experience roughly 9 dB less path loss than 2.4 GHz at the same distance, making them the default for long-range, low-power IoT. Frequency band selection requires balancing range, data rate, power budget, interference, and regulatory constraints.

Key Concepts

  • Multipath Propagation: Signal reaching receiver via multiple paths (direct, reflected, diffracted); causes constructive/destructive interference
  • Fading: Rapid signal variation due to multipath interference; Rayleigh fading for environments without line-of-sight
  • Shadowing: Slow signal variation due to large obstructions (buildings, hills); modeled as log-normal distribution with 6-12 dB std deviation
  • Path Loss Exponent: Constant n in path loss model P ∝ d^(-n); n=2 for free space, n=3-4 for indoor, n=4-6 for obstructed environments
  • Link Budget: System-level calculation: TX power + TX antenna gain - path loss - margins ≥ RX sensitivity
  • Fade Margin: Extra link budget reserved to maintain connectivity during deep fades; typically 10-20 dB for IoT
  • Wall Attenuation Factor: Additional loss per wall; drywall ~3 dB, concrete ~10-15 dB, reinforced concrete ~20+ dB
  • SINR (Signal-to-Interference-plus-Noise Ratio): Metric for link quality in the presence of interference; determines achievable data rate

11.1 Introduction

⏱️ ~8 min | ⭐⭐⭐ Advanced | 📋 P08.C16D.U01

Understanding wireless propagation is essential for predicting coverage, planning deployments, and troubleshooting connectivity issues. This chapter brings together the physics of path loss with practical design considerations, including interference mitigation and a framework for selecting the optimal frequency band for your IoT application.

Learning Objectives

By the end of this chapter, you will be able to:

  • Calculate free-space path loss (FSPL) for a given distance and frequency, and predict how doubling either parameter shifts the loss budget
  • Classify common interference sources by frequency band and assess their impact on IoT link reliability
  • Evaluate coexistence strategies (frequency hopping, channel selection, CSMA/CA, spread spectrum) for a given deployment scenario
  • Select an optimal wireless frequency band for an IoT deployment using a structured decision framework that weighs range, data rate, power, and regulatory constraints
  • Design a complete link budget for an IoT wireless solution, incorporating antenna gains, fade margins, and real-world path loss exponents

When radio waves travel from a transmitter to a receiver, they bounce off walls, pass through objects, and spread out over distance. This behavior is called propagation. Understanding propagation helps you predict wireless range and design IoT networks that work reliably in real buildings and outdoor environments.

“Radio signals are like flashlight beams,” explained Max the Microcontroller, shining a flashlight across the room. “The farther away you are, the dimmer the light gets. And if there are walls or furniture in the way, even less light reaches the other side.”

Sammy the Sensor held up his antenna. “That is why I sometimes lose connection to the gateway! The signal gets weaker as it travels.” Max nodded. “It is called path loss, and it follows a formula. Every time you double the distance, the signal drops by about 6 dB. And higher frequencies lose even more – a 2.4 GHz signal loses about 9 dB more than an 868 MHz signal over the same distance.”

“So that is why sub-GHz radios reach farther!” said Bella the Battery excitedly. “Lower frequency means less path loss, which means I do not need to boost my transmit power as much. More range with less energy!”

Lila the LED added a practical tip. “When designing an IoT network, you cannot just use the math. Real buildings have walls, metal ducts, and people moving around. You need to test your signal strength at different locations – that is called a site survey. Plan for the worst case, and your network will work reliably even on bad days!”

11.2 Prerequisites

Before diving into this chapter, you should be familiar with:

Series Navigation:

Hands-On:

Design Resources:

11.3 Wireless Propagation Characteristics

⏱️ ~18 min | ⭐⭐⭐ Advanced | 📋 P08.C16D.U02

How It Works: Free Space Path Loss

Radio signals weaken as they travel through space following an inverse square law. The Free Space Path Loss (FSPL) formula quantifies this: FSPL(dB) = 20log₁₀(d) + 20log₁₀(f) + 32.45, where distance and frequency both increase loss. Doubling the distance adds 6 dB loss. Moving from 868 MHz to 2.4 GHz adds approximately 9 dB loss at the same distance. This fundamental relationship explains why sub-GHz signals travel farther than higher frequencies.

11.3.1 Frequency vs Range Trade-off

The choice of frequency band involves fundamental trade-offs between range, bandwidth, and penetration:

Graph showing the inverse relationship between frequency and range for wireless IoT, with higher frequencies enabling more bandwidth but shorter range and greater obstacle sensitivity
Figure 11.1: Rule-of-thumb trade-off: higher frequency enables more bandwidth but reduces range/penetration and increases sensitivity to obstacles.

11.3.2 Free Space Path Loss

Signal strength decreases with distance according to the free space path loss formula:

\[ FSPL(dB) = 20\log_{10}(d) + 20\log_{10}(f) + 32.45 \]

Where: - \(d\) = distance in kilometers - \(f\) = frequency in MHz

Key insight: Path loss increases with both distance AND frequency. A 5 GHz signal experiences about 6.4 dB more path loss than a 2.4 GHz signal at the same distance (20 log10(5000/2400) = 6.4 dB).

Tradeoff: Directional Antenna vs Omnidirectional Antenna

Option A (Directional): Gain 8-24 dBi, beamwidth 15-60 degrees, effective range 2-10x omnidirectional, requires precise alignment. Example: 12 dBi Yagi at 868 MHz achieves 8 km point-to-point link with +14 dBm TX power.

Option B (Omnidirectional): Gain 2-6 dBi, beamwidth 360 degrees horizontal, range 0.5-3 km typical, no alignment needed. Example: 3 dBi dipole at 868 MHz covers 2 km radius with +14 dBm TX power.

Decision Factors: Choose directional for fixed point-to-point backhaul links, rural base stations serving distant sensors, or when interference rejection is critical (narrower beam rejects off-axis interference by 10-20 dB). Choose omnidirectional for mobile assets, star topology with sensors in all directions, urban gateways, or when installation simplicity outweighs range optimization.

Practical Example: Path Loss Comparison

For a device 10 meters away:

At 868 MHz (sub-GHz): \[FSPL = 20\log_{10}(0.01) + 20\log_{10}(868) + 32.45 = 51.2 \text{ dB}\]

At 2.4 GHz: \[FSPL = 20\log_{10}(0.01) + 20\log_{10}(2400) + 32.45 = 60.0 \text{ dB}\]

At 5 GHz: \[FSPL = 20\log_{10}(0.01) + 20\log_{10}(5000) + 32.45 = 66.4 \text{ dB}\]

The sub-GHz signal has 8.8 dB less path loss than 2.4 GHz, meaning it requires less transmit power or achieves greater range.

Determine the required transmit power at each frequency to achieve the same received signal strength (-60 dBm) at 10 meters:

\[P_{TX} = P_{RX} + \text{FSPL}\]

868 MHz: \(P_{TX} = -60 + 51.2 = -8.8\) dBm (0.13 mW)

2.4 GHz: \(P_{TX} = -60 + 60.0 = 0\) dBm (1 mW)

5 GHz: \(P_{TX} = -60 + 66.4 = 6.4\) dBm (4.4 mW)

Sub-GHz requires \(7.5\times\) less power than 2.4 GHz, and \(33\times\) less than 5 GHz for the same received power.

11.4 Interactive: Free-Space Path Loss Calculator

11.5 Interference and Coexistence

⏱️ ~10 min | ⭐⭐ Intermediate | 📋 P08.C16D.U03

11.5.1 Sources of Interference

Understanding potential interference sources helps in selecting the appropriate frequency band:

Chart of common interference sources by frequency band showing 2.4 GHz congested by Wi-Fi, Bluetooth, microwaves; 5 GHz cleaner with DFS radar; sub-GHz with different regional interferers
Figure 11.2: Common interference sources by band: 2.4 GHz is usually busiest; 5 GHz is typically cleaner (but may require DFS); sub‑GHz has different local interferers.

11.5.2 Coexistence Strategies

IoT protocols employ various techniques to coexist in crowded spectrum:

  1. Frequency Hopping (Bluetooth): Rapidly switches between channels
  2. Channel Selection (Wi-Fi): Chooses less congested channels
  3. CSMA/CA (Wi-Fi, Zigbee): Listen before transmit
  4. Spread Spectrum (LoRa): Spreads signal across wide bandwidth
  5. Time Division (WirelessHART): Allocates specific time slots

11.5.3 Why Real-World Path Loss Is 2-3x Worse Than Free Space (and How to Budget for It)

The FSPL formula predicts loss in a perfect vacuum with no obstacles. In practice, IoT deployments experience 2 to 3 times more loss than FSPL predicts, and understanding why is essential for reliable link budgets.

The path loss exponent tells the story. FSPL assumes a path loss exponent of 2 (signal decays with the square of distance). Real environments have higher exponents:

Environment Path Loss Exponent (n) Example at 1 km, 868 MHz
Free space 2.0 91 dB (FSPL)
Rural farmland 2.5-2.8 114-125 dB
Suburban residential 3.0-3.5 137-160 dB
Dense urban 3.5-4.5 160-205 dB
Indoor office (same floor) 3.0-4.0 137-182 dB
Indoor-to-outdoor 4.0-6.0 182-273 dB

Three physical mechanisms cause the extra loss:

  1. Multipath fading. Signals reflect off buildings, ground, and vehicles, arriving at the receiver as multiple copies with different phases. When copies cancel each other (destructive interference), signal strength drops by 20-40 dB in a matter of centimeters. This is why a sensor that works on a desk may fail when moved 10 cm sideways.

  2. Fresnel zone obstruction. Even without touching obstacles, a signal weakens when objects intrude into its Fresnel zone – the football-shaped region around the line-of-sight path. At 868 MHz over 1 km, the first Fresnel zone has a 9.3 m radius at midpoint. A tree canopy blocking 60% of this zone adds 6-10 dB of loss that FSPL does not predict. This is why agricultural sensors work well in winter (bare trees) but fail in summer (full canopy).

  3. Material attenuation. Every wall, window, and floor the signal passes through absorbs energy:

    • Drywall: 3-4 dB per wall
    • Concrete: 10-15 dB per wall
    • Reinforced concrete: 20-25 dB per wall
    • Low-E glass (energy-efficient windows): 25-30 dB (the metallic coating acts as a Faraday cage)

Practical link budget rule of thumb: Add a 20 dB fade margin to your FSPL calculation. This covers typical multipath fading (10 dB for 90th percentile) plus one concrete wall (10-15 dB). If your calculated link budget is exactly zero after adding fade margin, the link will fail roughly 50% of the time. A positive margin of 10+ dB gives reliable operation.

11.6 Design Considerations for IoT

⏱️ ~15 min | ⭐⭐⭐ Advanced | 📋 P08.C16D.U04

11.6.1 Frequency Band Selection Framework

Decision flowchart for choosing an IoT wireless frequency band based on range requirements, data rate needs, power constraints, and deployment environment
Figure 11.3: Decision flow for choosing an IoT band based on range, data rate, and power constraints.

11.6.2 Key Selection Criteria

When choosing a wireless frequency band for your IoT application, consider:

  1. Range Requirements
    • Indoor: 2.4 GHz or 5 GHz typically sufficient
    • Outdoor urban: 2.4 GHz or sub-GHz
    • Rural/wide area: Sub-GHz or cellular
  2. Data Rate Needs
    • Video streaming: 5 GHz Wi-Fi
    • Sensor data: 2.4 GHz mesh or sub-GHz
    • Occasional updates: Sub-GHz LPWAN
  3. Power Constraints
    • Battery-powered, multi-year: Sub-GHz LPWAN
    • Mains-powered: Any band suitable
    • Coin cell: BLE or LPWAN
  4. Deployment Environment
    • Dense urban: Expect 2.4 GHz congestion
    • Industrial: Consider sub-GHz for penetration
    • Residential: 2.4 GHz or 5 GHz viable
  5. Regulatory Compliance
    • Check regional spectrum allocation
    • Verify power limits
    • Confirm duty cycle restrictions

11.8 Practice Exercises

  1. Pick a target deployment (smart building, agriculture, asset tracking) and write down: range, payload size, message frequency, and battery target.
  2. Compute wavelengths for 868/915 MHz, 2.4 GHz, and 5 GHz using λ = c / f. Write one sentence on what that implies for obstacle penetration.
  3. Use FSPL to compare 868 MHz vs 2.4 GHz at the same distance: what is the dB penalty at 1 km? (Hint: FSPL includes a 20·log10(f) term.)
  4. Choose a band and justify it in two bullets: one for physics (path loss/penetration) and one for operations (interference, regulations, deployment complexity).
  5. If you plan to ship globally, list which regions require different sub‑GHz configs (EU868 vs US915 vs AS923) and how you’d expose that choice (SKU vs firmware).

For hands-on spectrum work, continue with Mobile Wireless Labs and Implementation.

Scenario: You need to place a LoRa gateway to cover agricultural sensors spread across a 3 km radius. Sensors use SF12 (spreading factor 12) at 868 MHz with +14 dBm TX power and 2 dBi omnidirectional antennas. Will they reach the gateway?

Step 1: Calculate Free Space Path Loss (FSPL)

FSPL = 20·log₁₀(d_km) + 20·log₁₀(f_MHz) + 32.45
     = 20·log(3) + 20·log(868) + 32.45
     = 9.54 + 59.16 + 32.45
     = 101.15 dB

Step 2: Add environmental losses

Rural farmland (light vegetation, undulating terrain):
  Foliage attenuation:        5 dB (spring/summer)
  Ground reflection loss:     3 dB
  Fresnel zone obstruction:   2 dB (trees at 40% of first Fresnel zone)
  Total environmental loss:  10 dB

Step 3: Calculate received signal strength

TX power:           +14 dBm
TX antenna gain:     +2 dBi
FSPL:              -101.15 dB
Environmental:      -10 dB
RX antenna gain:     +6 dBi (directional Yagi at gateway)
──────────────────────────────
Received power:     -89.15 dBm

Step 4: Compare to receiver sensitivity

LoRa receiver (SF12, 125 kHz bandwidth): -137 dBm sensitivity

Link margin = Received power - Sensitivity
            = -89.15 - (-137)
            = +47.85 dB

Step 5: Evaluate link quality

Typical link margin targets:
  10-15 dB: Marginal (works in ideal conditions, fails in rain/storms)
  15-20 dB: Acceptable (reliable most of the time)
  20-30 dB: Good (works through moderate fading)
  >30 dB:   Excellent (very reliable, tolerates heavy rain/foliage)

This link: 47.85 dB = Excellent

Result: A single gateway at 3 km with a 6 dBi antenna provides robust coverage even through heavy rain and full foliage. The 47 dB margin allows reducing TX power to +8 dBm (saving 75% battery energy) while maintaining 41 dB margin, or extending range to 6-8 km in favorable terrain.

Requirement Indoor (Office/Home) Indoor (Industrial) Outdoor (Urban) Outdoor (Rural)
Typical range needed 20-50m 50-200m 500m-2km 2-15km
Primary obstacle Drywall (3-4 dB/wall) Concrete/metal (10-25 dB/wall) Buildings (15-30 dB) Foliage, terrain
Recommended frequency 2.4 GHz or 5 GHz Sub-GHz (868/915 MHz) Sub-GHz or 2.4 GHz Sub-GHz LoRaWAN
Typical technology Wi-Fi, BLE, Zigbee Sub-GHz proprietary, WirelessHART Zigbee/Thread mesh, LoRaWAN LoRaWAN, NB-IoT
Path loss example (50m) 60-70 dB (2.4 GHz) 80-100 dB (2.4 GHz through metal) 85-110 dB (NLOS urban) 100-120 dB (rolling terrain)
Key design challenge Multipath fading from furniture/metal Penetration through industrial equipment Non-line-of-sight around buildings Fresnel zone clearance over terrain
Power budget Moderate (can use mains or battery) Mix (mains for routers, battery for endpoints) Battery (solar for gateways) Battery (multi-year required)

Material attenuation quick reference (2.4 GHz):

  • Drywall: 3-4 dB
  • Brick wall: 5-10 dB
  • Concrete wall: 10-15 dB
  • Reinforced concrete: 20-25 dB
  • Metal stud wall: 8-12 dB
  • Floor (between levels): 12-20 dB
  • Low-E glass window: 25-30 dB (metallic coating acts as Faraday cage)

Sub-GHz advantage: Same materials attenuate ~3-5 dB less at 868 MHz vs 2.4 GHz due to longer wavelength diffraction.

Common Mistake: Using Free Space Path Loss for Indoor Range Calculations

The Error: An engineer calculates FSPL for a 50-meter indoor deployment at 2.4 GHz and gets 67 dB. The link budget shows 30 dB margin, so they confidently deploy 100 sensors expecting reliable coverage.

Free Space Path Loss calculation (WRONG for indoor):

FSPL(50m, 2.4 GHz) = 20·log(0.05) + 20·log(2400) + 32.45
                    = -26.02 + 67.60 + 32.45
                    = 74.03 dB

Why it fails in a real building:

Actual path: Sensor in Meeting Room B → Through 2 drywall walls → Across open office → Through metal-stud wall → Router in IT closet

Path loss breakdown:
  FSPL (50m):                74 dB (baseline)
  Drywall walls (2 × 4 dB):   8 dB
  Metal-stud wall:           12 dB
  Multipath fading (Rayleigh): 10-20 dB (varies by location)
  Fresnel zone obstruction:   5 dB (desks/cubicles)
  Human body blockage:        3 dB (intermittent)
──────────────────────────────────────
  Total path loss:          112-122 dB (38-48 dB worse than FSPL)

Link budget reality:

TX power:        +20 dBm (Wi-Fi)
TX antenna:       +2 dBi
Path loss:      -112 to -122 dB
RX antenna:       +2 dBi
─────────────────────────────
Received:        -88 to -98 dBm

Wi-Fi sensitivity (54 Mbps): -70 dBm
Link margin: -98 - (-70) = -28 dB FAIL (signal 28 dB too weak)

Measured result:

  • 40% of sensors: Frequent disconnections
  • 25% of sensors: Cannot associate at all
  • Throughput drops to 1-6 Mbps (lowest modulation rates)
  • Retransmissions consume 5× more battery energy

The correct approach: Use empirical path loss models for indoor environments:

ITU indoor office model:

PL = 20·log(f) + N·log(d) + Lf(n) - 28
where:
  f = frequency (MHz)
  d = distance (m)
  N = path loss exponent (28-40, depends on building construction)
  Lf(n) = floor penetration loss (15-20 dB per floor)

For office with drywall (N=30):
PL(50m) = 20·log(2400) + 30·log(50) + 0 - 28
        = 67.6 + 50.97 - 28
        = 90.6 dB (16.6 dB worse than FSPL)

For office with concrete (N=35):
PL(50m) = 67.6 + 59.47 - 28 = 99.1 dB (25 dB worse than FSPL)

Best practice: Always add 20 dB fade margin to FSPL calculations for indoor deployments to account for multipath, obstacles, and human movement. Better yet, do a site survey with actual hardware before finalizing AP/gateway placement.

Objective: Calculate maximum range for a custom IoT deployment using real-world path loss models.

Setup: Open a spreadsheet or Python notebook.

Steps:

  1. Define your system parameters:
    • TX power: 14 dBm (LoRaWAN typical)
    • RX sensitivity: -137 dBm (SF12, 125 kHz BW)
    • Antenna gains: 2 dBi (node), 8 dBi (gateway)
    • Frequency: 868 MHz
    • Fade margin: 20 dB
  2. Calculate link budget: Budget = TX + TX_gain + RX_gain - RX_sens - Fade = 14 + 2 + 8 - (-137) - 20 = 141 dB
  3. For different environments, solve for distance d:
    • Rural (n=2.5): FSPL = 20log₁₀(d) + 20log₁₀(868) + 32.45 = 141 dB → d ≈ 15 km
    • Suburban (n=3.5): Effective FSPL = 141 dB, but use n=3.5 model → d ≈ 6 km
    • Urban (n=4.5): d ≈ 2.5 km
  4. Experiment: What if you use 433 MHz instead? (Hint: 20log₁₀(433/868) = -6 dB gain → ~1.5x longer range)

What to Observe: How small changes in frequency, antenna gain, or environment drastically change coverage. Try real deployment: what fade margin do you need for 99% reliability vs 90%?

11.9 Concept Relationships

Concept Relationship to Other Concepts Practical Impact
Free Space Path Loss Increases with both distance and frequency Determines maximum range for given TX power
Path Loss Exponent Modifies FSPL for real environments (n=2 to 6) Explains 2-3x worse real-world vs theory
Fade Margin Adds safety buffer to link budget 20 dB typical for 90% reliability
Fresnel Zone Clearance needed around line-of-sight path 60% clearance prevents 6-10 dB loss
Antenna Gain Directional focuses energy, omnidirectional spreads Directional adds 8-24 dBi for point-to-point
Sub-GHz Advantage ~9 dB less path loss than 2.4 GHz Enables 2-3x longer range or 8x less power

Common Pitfalls

Standard path loss models (ITU, Okumura-Hata) are statistical averages. Actual propagation in a specific building can deviate by 15-20 dB. Always calibrate the path loss model with site measurements before using it for coverage planning.

A link budget that exactly meets receiver sensitivity at the edge of coverage will be unreliable 50% of the time due to fading. Add 10-20 dB of fade margin depending on fading severity. Links without fade margin fail consistently in real deployments.

Indoor path loss exponents (3-4) are significantly higher than outdoor free-space (2). A sensor 50 meters away inside a building with three concrete walls may have path loss equivalent to 500 meters outdoors. Use indoor-specific propagation models for building deployments.

The human body causes 3-5 dB attenuation at 2.4 GHz and higher for body-worn devices. Wearable IoT sensors must account for on-body path loss in their link budget. Testing without a person holding or wearing the device produces overly optimistic results.

11.10 Summary

This chapter covered wireless propagation characteristics and design considerations:

  • Free space path loss increases with both distance and frequency, making sub-GHz bands optimal for long-range applications
  • Path loss comparison: Sub-GHz has ~9 dB less loss than 2.4 GHz at the same distance
  • Interference varies by band: 2.4 GHz is most congested; sub-GHz typically has minimal interference
  • Coexistence strategies include frequency hopping, channel selection, CSMA/CA, and spread spectrum
  • Band selection framework: Start with range requirements, then consider data rate and power constraints
  • Regulatory compliance varies by region; design firmware to support region-specific configurations

11.11 What’s Next

Chapter Focus
Mobile Wireless Labs and Implementation Hands-on experiments with RF spectrum analysis, channel scanning, and link budget calculations
Mobile Wireless Comprehensive Review Scenario-based recap of spectrum choices and design trade-offs
Wi-Fi Fundamentals and Standards Deep dive into 802.11 protocols, channel bonding, and MIMO techniques
Bluetooth Overview Bluetooth Classic and BLE coexistence, adaptive frequency hopping, and power profiles

11.12 See Also

11.13 Knowledge Check

Scenario: You’re deploying 200 soil moisture, temperature, and rainfall sensors across a 10 km² farm with rolling hills (50–100 m elevation changes), scattered barns, dense tree groves, and no existing power/network infrastructure. Sensors transmit ~100‑byte readings every 15 minutes. Requirements: 5+ year battery life on 2× AA lithium cells and a single gateway covering the entire farm.

Think about:

  1. How does free-space path loss scale with frequency and distance?
  2. Why are lower frequencies often more forgiving with foliage and non-line-of-sight paths (all else equal)?
  3. What changes in the link budget matter most here: transmit power, antenna gains, receiver sensitivity, airtime, or topology?

Key Insight: For sparse rural deployments, sub‑GHz LPWANs are often a strong default because they offer high link budgets at low data rates. Range still depends on antenna heights, terrain, foliage, device settings, and local regulatory limits.

FSPL quick comparison (free space):

FSPL(dB) = 32.44 + 20·log₁₀(d_km) + 20·log₁₀(f_MHz)

At d = 10 km:
868 MHz  → ~111 dB
2.4 GHz  → ~120 dB (≈9 dB higher)
5.0 GHz  → ~126 dB (≈15 dB higher)

That ≈9 dB penalty corresponds to roughly 8× more transmit power (or ~9 dB more link margin) to achieve the same received power in free space.

Why sub‑GHz often wins here:

  • Coverage: Multi‑kilometer links are feasible when the link budget is high and antenna placement is good.
  • Penetration/terrain: Lower frequencies are generally more forgiving with foliage and non-line-of-sight paths.
  • Power: Infrequent transmissions plus a high link budget can enable multi‑year batteries.
  • Topology: A single gateway is much easier to operate than a dense mesh of powered routers.

Battery-life sanity check (not a guarantee):

  • Estimate airtime per message (depends on PHY/data rate) and multiply by messages/day.
  • Compute average current: I_avg ≈ (I_tx·t_tx + I_rx·t_rx + I_sleep·t_sleep) / 24h
  • Compare to battery capacity (and include temperature, self‑discharge, retries, and battery aging).

Verify Your Understanding:

  • Compute the FSPL difference between 868 MHz and 2.4 GHz at the same distance. How does that translate into link-budget requirements?
  • Which two deployment choices (gateway height, antenna choice, payload frequency, etc.) would most improve reliability without increasing transmit power?

Scenario: Your startup develops a LoRa asset tracker for shipping containers traveling globally (US→Europe→Asia). Product spec: a single hardware SKU with firmware-selectable regional settings. Engineering proposes shipping everything with a US915 configuration “to keep it simple.”

Think about:

  1. What changes across regions besides the center frequency (channel plan, max EIRP, duty-cycle/LBT, certification requirements)?
  2. Why do different regions allocate different unlicensed sub‑GHz bands (and why can rules change over time)?
  3. What provisioning workflow prevents misconfiguration (factory programming, installer app, geo-locked config, region SKUs)?

Key Insight: There is no single global sub‑GHz configuration. Regional allocations, power limits, and channel plans differ; using the wrong regional parameters can cause harmful interference and fail regulatory/certification testing.

Illustrative regional differences (always confirm locally):

Region          Common SRD/ISM bands     Typical constraints (examples)
━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
North America   902–928 MHz              Different channel plan; constraints differ from EU
Europe          863–870 MHz              Lower EIRP; duty-cycle or LBT in many sub-bands
Japan           920–928 MHz              Region-specific LBT and channelization
China           470–510 MHz (or others)  Region-specific allocations and certification

Design pattern: region-aware firmware

  • Implement region profiles matching your protocol’s official regional parameters (e.g., LoRaWAN EU868, US915, AU915, AS923, CN470, IN865, …).
  • Store the selected region in non-volatile storage and expose it in a diagnostic screen/log.
  • Prevent accidental misconfiguration (e.g., “EU device set to US915”) with installer tooling and clear labeling.
// Multi-region firmware configuration
typedef enum {
    REGION_EU868,    // Europe: 863-870 MHz
    REGION_US915,    // North America: 902-928 MHz
    REGION_AU915,    // Australia: 915-928 MHz
    REGION_AS923,    // Asia: 920-923 MHz
    REGION_CN470,    // China: 470-510 MHz
    REGION_IN865     // India: 865-867 MHz
} LoRaRegion_t;

void configureRegion(LoRaRegion_t region) {
    switch(region) {
        case REGION_EU868:
            setFrequencyBand(863000000, 870000000);
            setMaxTxPower(14);  // Example: follow regional parameters
            setDutyCycle(1);    // Example: duty-cycle/LBT policy is region-specific
            break;
        case REGION_US915:
            setFrequencyBand(902000000, 928000000);
            setMaxTxPower(30);  // Example: follow regional parameters
            setDutyCycle(0);    // Example: constraints differ by region (not "no rules")
            break;
        // ... other regions
    }
}

Verify Your Understanding:

  • Design a firmware architecture that supports region selection without requiring different hardware SKUs
  • List 2–3 product decisions that reduce misconfiguration risk (labeling, installer workflow, geo-locking, QA checks)