1072 LPWAN Assessment: Technology Selection
1072.1 Introduction
This chapter provides comprehensive assessment scenarios for LPWAN technology selection, including detailed analysis of real-world deployment decisions involving cost, scale, and technical constraints.
- Apply multi-criteria decision making for LPWAN technology selection
- Analyze complex deployment scenarios with competing constraints
- Calculate and compare Total Cost of Ownership across technologies
- Understand spectrum allocation differences between LPWAN technologies
1072.2 Technology Selection Scenario
Complex deployment decisions require careful analysis of all constraints.
Question: A logistics company wants to deploy 10,000 asset trackers that report GPS location + temperature every 15 minutes (96 messages/day). Each message payload is 45 bytes. Battery life must exceed 3 years. They’re evaluating LoRaWAN vs Sigfox. Which technology should they choose and why?
This demonstrates LPWAN technology selection based on application constraints:
From the text - LoRaWAN vs Sigfox:
LoRaWAN: - Data rate: 0.3-50 kbps (adaptive) - Max Payload: 243 bytes - Messages/Day: Unlimited (duty cycle limited) - Battery Life: 5-10 years
Sigfox: - Data rate: 100 bps uplink, 600 bps downlink - Max Payload: 12 bytes (uplink) - Messages/Day: 140 uplink, 4 downlink (hard limit) - Battery Life: 10-20 years
Application Requirements vs Technology Capabilities:
| Requirement | Value | LoRaWAN | Sigfox | Result |
|---|---|---|---|---|
| Messages/day | 96 | Unlimited | Max 140 | Sigfox barely OK |
| Payload size | 45 bytes | 243 max | 12 max | Sigfox FAILS |
| Battery life | 3+ years | 5-10 yr | 10-20 yr | Both OK |
| Device count | 10,000 | Scalable | Scalable | Both OK |
| Bi-directional | Yes (GPS) | A/B/C | Limited | LoRaWAN better |
Critical Failure Analysis:
The 45-byte payload requirement for GPS tracker data cannot fit in Sigfox’s 12-byte maximum:
| Field | Bytes | Purpose |
|---|---|---|
| GPS Lat | 4 | Latitude coordinate |
| GPS Lon | 4 | Longitude coordinate |
| Altitude | 2 | Height above sea level |
| Speed | 2 | Movement speed |
| Heading | 2 | Direction of travel |
| Temperature | 2 | Sensor reading |
| Battery | 1 | Remaining charge |
| Device ID | 4 | Identification |
| Timestamp | 4 | Reading time |
| Status Flags | 2 | Device status |
| Checksum | 2 | Data integrity |
| Total | 29+ bytes | Minimum GPS payload |
Even with extreme compression, the fundamental data requirements exceed Sigfox’s 12-byte limit.
Battery Life Calculation (SF7, 45-byte payload, 96 msg/day):
| Component | Value | Calculation |
|---|---|---|
| TX current | 120 mA | During transmission |
| TX time | 200 ms | 45 bytes at SF7 |
| TX per day | 96 | Every 15 minutes |
| TX energy/day | 0.64 mAh | (120 mA x 0.2s / 3600s) x 96 |
| Sleep current | 5 uA | Deep sleep mode |
| Sleep energy/day | 0.12 mAh | (5 uA / 1000) x 24h |
| Total daily | 0.76 mAh/day | TX + Sleep |
| Battery capacity | 2400 mAh | 2x AA lithium batteries |
| Battery life | 8.6 years | 2400 / (0.76 x 365) |
| vs Requirement | 2.9x margin | Exceeds 3-year requirement |
Summary - Why LoRaWAN is the only choice:
- Payload: 45 bytes fits in LoRaWAN (243 max) but exceeds Sigfox (12 max) - DEAL BREAKER
- Messages: 96/day well within LoRaWAN capacity, marginal on Sigfox
- Battery: Both exceed 3-year requirement
- Flexibility: LoRaWAN allows future expansion, Sigfox is maxed out
- Downlinks: LoRaWAN unlimited, Sigfox only 4/day
The payload constraint alone eliminates Sigfox, making LoRaWAN the only viable LPWAN option for this application.
1072.3 Large-Scale Cost Analysis
Understanding TCO at scale is critical for enterprise deployments.
Question: A water utility plans to deploy 50,000 smart water meters across a region. Each meter sends one reading per day (24 bytes). Comparing 5-year TCO: Private LoRaWAN costs 1,500 EUR/gateway x 30 gateways + 15 EUR/sensor + 300 EUR/month network server. NB-IoT costs 20 EUR/sensor + 1.50 EUR/device/month data plan. What is the approximate cost difference favoring the cheaper option?
This demonstrates LPWAN TCO analysis for large-scale deployments:
Complete Cost Breakdown:
Private LoRaWAN costs
Initial investment (Year 1):
Component Calculation Cost Gateways 30 x 1,500 EUR 45,000 EUR Sensors 50,000 x 15 EUR 750,000 EUR Installation (estimate) - 100,000 EUR Total initial 895,000 EUR Recurring costs (Years 1-5):
Component Calculation Cost Network server 300 EUR/month x 12 x 5 18,000 EUR Maintenance 5,000 EUR/year x 5 25,000 EUR Total 5-year recurring 43,000 EUR
5-year total cost of ownership (TCO): 895,000 + 43,000 = 938,000 EUR.
NB-IoT (cellular) costs
Initial investment (Year 1):
Component Calculation Cost Sensors 50,000 x 20 EUR 1,000,000 EUR Installation (estimate) - 100,000 EUR Total initial 1,100,000 EUR Recurring costs (Years 1-5):
Component Calculation Cost Data plan 50,000 devices x 1.50 EUR/month x 12 x 5 4,500,000 EUR
5-year total cost of ownership (TCO): 1,100,000 + 4,500,000 = 5,600,000 EUR.
Cost comparison
| Technology | 5-year TCO | Breakdown |
|---|---|---|
| Private LoRaWAN | 938,000 EUR | 895k initial + 43k recurring |
| NB-IoT cellular | 5,600,000 EUR | 1.1M initial + 4.5M recurring |
Difference: 5,600,000 - 938,000 = 4,662,000 EUR - NB-IoT costs significantly more.
Year-by-Year Cost Analysis (Cumulative TCO):
| Year | LoRaWAN TCO | NB-IoT TCO | Difference | LoRaWAN Savings |
|---|---|---|---|---|
| 1 | 903,600 EUR | 2,000,000 EUR | 1,096,400 EUR | 54% cheaper |
| 2 | 912,200 EUR | 2,900,000 EUR | 1,987,800 EUR | 69% cheaper |
| 3 | 920,800 EUR | 3,800,000 EUR | 2,879,200 EUR | 76% cheaper |
| 4 | 929,400 EUR | 4,700,000 EUR | 3,770,600 EUR | 80% cheaper |
| 5 | 938,000 EUR | 5,600,000 EUR | 4,662,000 EUR | 83% cheaper |
Key Insight: LoRaWAN’s cost advantage grows over time due to minimal recurring costs (8.6k EUR/year) vs NB-IoT’s massive subscriptions (900k EUR/year).
Cost Per Device Analysis (50,000 devices):
| Metric | LoRaWAN | NB-IoT | Ratio |
|---|---|---|---|
| Total 5-year TCO | 938,000 EUR | 5,600,000 EUR | 6.0x |
| Cost per device (5yr) | 18.76 EUR | 112.00 EUR | 6.0x |
| Cost per device per year | 3.75 EUR | 22.40 EUR | 6.0x |
| Cost per device per month | 0.31 EUR | 1.87 EUR | 6.0x |
Conclusion: NB-IoT costs 6x more per device due to recurring subscription fees.
Why the Difference is So Large:
| Cost Type | LoRaWAN (5yr) | NB-IoT (5yr) | Difference | % of Total Diff |
|---|---|---|---|---|
| Initial hardware | 895,000 EUR | 1,100,000 EUR | 205,000 EUR (NB-IoT more) | 4% |
| Recurring (5yr) | 43,000 EUR | 4,500,000 EUR | 4,457,000 EUR | 96% |
| Total TCO | 938,000 EUR | 5,600,000 EUR | 4,662,000 EUR | 100% |
Analysis: 96% of cost difference comes from recurring subscription fees. The larger the scale and longer the timeframe, the more dominant LoRaWAN’s cost advantage becomes.
1072.4 Spectrum Allocation Understanding
Question: An IoT architect must explain LPWAN spectrum allocation to stakeholders. Which statement is MOST accurate regarding LPWAN spectrum usage?
Option B is correct - LPWAN technologies split between spectrum types:
Unlicensed ISM Bands (LoRaWAN, Sigfox): - Regions: - EU: 868 MHz (863-870 MHz) - US: 915 MHz (902-928 MHz) - Asia: 923 MHz (920-925 MHz) - Advantages: - No spectrum licensing fees - No recurring regulatory costs - Freely deployable private networks - Disadvantages: - Duty cycle restrictions (1% EU, listen-before-talk US) - Interference from other ISM users (industrial equipment, home automation) - No QoS guarantees - Power limits (14 dBm EU, 30 dBm US)
Licensed Cellular Spectrum (NB-IoT, LTE-M): - Bands: LTE bands (700, 800, 900, 1800, 2100, 2600 MHz) - Advantages: - Protected spectrum (no interference) - Guaranteed QoS and SLAs - Higher allowed transmit power (23 dBm) - No duty cycle restrictions - Disadvantages: - Requires cellular subscription (2-10 EUR/device/month) - Carrier-dependent (no private network option) - Spectrum auction costs passed to consumers
Why A is wrong: NB-IoT and LTE-M do NOT use ISM bands.
Why C is wrong: LoRaWAN and Sigfox use unlicensed, not licensed.
Why D is wrong: Spectrum licensing has no inherent relationship to coverage - it depends on frequency (lower is generally better), power, and modulation.
Strategic implication: Choose unlicensed (LoRaWAN) for private networks with zero recurring costs, or licensed (NB-IoT) for carrier-managed networks with guaranteed QoS.
1072.5 Agricultural Deployment Scenario
Question: A company evaluates LPWAN options for 10,000 remote agricultural sensors across 200 sq km. Each sensor reports soil data (50 bytes) twice daily for 10 years. Cellular coverage exists but is spotty. They can install infrastructure on water towers and grain silos. What is the MOST cost-effective strategy?
Private LoRaWAN (C) is most cost-effective at this scale:
Cost Analysis:
LoRaWAN (Private): - Sensors: 10,000 x $15 = $150,000 - Gateways: 200 sq km / 4 sq km per gateway = 50 gateways x $1,500 = $75,000 - Network server: $5,000/year x 10 years = $50,000 - Total: $275,000 over 10 years
Sigfox (Operator): - Sensors: 10,000 x $10 = $100,000 - Subscription: 10,000 x $6/year x 10 years = $600,000 - Total: $700,000 (2.5x more than LoRaWAN)
NB-IoT (Cellular): - Sensors: 10,000 x $20 = $200,000 - Subscription: 10,000 x $24/year x 10 years = $2,400,000 - Total: $2,600,000 (9.5x more than LoRaWAN!)
Why LoRaWAN wins: 1. Infrastructure control - Company owns water towers and grain silos (perfect gateway locations) 2. Zero recurring costs - No subscriptions after initial deployment 3. Scale economics - At 10,000 devices, gateway cost ($75k) amortizes to $7.50/device 4. Coverage - 200 sq km rural area well-suited for LoRa’s 15km range 5. 10-year lifespan - Private network costs are upfront; cellular costs compound annually
Hybrid approach (D) would cost more than pure LoRaWAN while adding complexity. Since they can install gateways on existing structures, achieving 100% LoRaWAN coverage is feasible.
ROI calculation: LoRaWAN saves $425k-$2.3M over alternatives. Payback period: Immediate (no recurring costs vs. $60k-$240k/year for alternatives).
1072.6 Summary
This technology selection assessment covered:
- Complex Deployment Scenarios: Analyzing GPS tracker requirements against LoRaWAN and Sigfox constraints
- Enterprise TCO Analysis: Understanding cost dynamics for 50,000+ device deployments
- Spectrum Considerations: Licensed vs unlicensed spectrum trade-offs
- Agricultural Use Cases: Optimizing for long-term, large-scale rural deployments
Key Takeaways: - Payload constraints can eliminate technologies regardless of other advantages - Recurring subscription costs dominate cellular IoT TCO at scale - Private networks become economical at ~1,000+ devices with 5+ year deployments - Infrastructure availability (towers, silos) enables cost-effective private network coverage
1072.7 What’s Next
Continue your LPWAN assessment with:
- LPWAN Assessment: Regulatory Compliance - Duty cycle, scaling, and operator risk scenarios
- LPWAN Assessment: Fundamentals - Review foundational concepts
- LoRaWAN Architecture - Deep dive into LoRaWAN network components
- NB-IoT Fundamentals - Cellular IoT technology details