PAN (Personal Area Network): A network spanning a person’s immediate surroundings (< 10 m); Bluetooth, ZigBee, and NFC operate at PAN scale
LAN (Local Area Network): A network within a building or campus (< 1 km); Wi-Fi and Ethernet are the dominant LAN technologies
WAN (Wide Area Network): A network spanning cities or countries; cellular networks and the internet are WANs
LPWAN (Low-Power Wide Area Network): A network class optimised for battery-powered IoT devices needing kilometre-scale range; LoRaWAN, NB-IoT, Sigfox
BAN (Body Area Network): A network on or near the human body; used for wearable health monitors
NAN (Neighbourhood Area Network): A network at smart grid or smart city scale (neighbourhood level), typically using mesh protocols
Classification Dimensions: Networks are classified by geographic scale, data rate, power consumption, and number of devices supported
62.1 In 60 Seconds
Networks are classified by coverage area: PAN (1-100m, Bluetooth/Zigbee), LAN (100m-1km, Wi-Fi/Ethernet), and WAN (>1km, LoRaWAN/cellular). Each classification implies trade-offs between bandwidth and range – PANs offer low power but limited throughput, while WANs provide kilometers of coverage at the cost of higher latency and data rates. Matching the right network class to your IoT deployment’s range, data rate, and power requirements is the first step in protocol selection.
Learning Objectives
By the end of this section, you will be able to:
Classify IoT protocols into their correct network tiers (PAN, LAN, WAN) based on range, data rate, and power constraints
Analyze bandwidth and coverage trade-offs to justify protocol selection for a given IoT deployment
Design multi-tier IoT network topologies that combine PAN, LAN, and WAN technologies appropriately
Select the optimal network classification and specific protocol based on measurable deployment requirements
Differentiate common IoT networking terminology and explain how each concept applies to real-world deployments
62.2 Prerequisites
Before diving into this chapter, you should be familiar with:
Networks are classified by how far they reach - like how we describe distances:
Walking distance = PAN (Personal Area Network) - devices within arm’s reach (1-100 meters)
Driving distance = LAN (Local Area Network) - devices in a building or campus (100m-1km)
Flying distance = WAN (Wide Area Network) - devices across a city or country (1km to global)
Network Type
Range
Example
PAN
1-100m
Smartwatch connecting to phone
LAN
100m-1km
Office Wi-Fi network
WAN
>1km
City-wide sensor network
Sensor Squad: Finding the Right Size Network!
“Not all networks are the same size,” said Max the Microcontroller. “Think of it like distances: walking distance, driving distance, and flying distance. Each needs a different type of connection.”
“I am a fitness tracker, so I use a PAN – a Personal Area Network,” said Sammy the Sensor. “Bluetooth connects me to the phone just a meter away. Short range, but super low power. My battery lasts a week!”
Lila the LED was in a smart building. “I am on a LAN – a Local Area Network. Wi-Fi covers the whole building, about 100 meters range. Plenty of bandwidth for all the lights, thermostats, and security cameras in the office.”
“And I am a soil moisture sensor on a farm,” added Bella the Battery. “I need a WAN – a Wide Area Network. LoRaWAN can reach the gateway 10 kilometers away! But the trade-off is speed – I can only send tiny messages. The rule of thumb: longer range means lower data rate. You cannot have both long range AND high speed without using lots of power. Pick the network classification that matches YOUR needs: range, data rate, and power budget!”
62.3 Bandwidth and Coverage Trade-offs
Time: ~6 min | Difficulty: Intermediate | Reference: P07.C11.U07
Today’s IoT networks are best explained by looking at the bandwidth and coverage of each network technology. Different protocols occupy different positions in the bandwidth-coverage space.
Figure 62.1: Bandwidth versus coverage quadrant chart for IoT protocols
Bandwidth-Coverage Analysis
Quadrant 1 (High Bandwidth, High Coverage): Ideal but expensive - 5G Cellular: Ultimate performance, high cost - Use cases: Autonomous vehicles, industrial automation, video surveillance
Putting Numbers to It
These bandwidth-coverage quadrants have quantifiable boundaries. The fundamental trade-off comes from the Shannon-Hartley theorem and path loss physics.
For a 100 mW transmitter at 900 MHz, the received power at distance \(d\) meters follows the Friis equation:
For example, at 1 km distance with 0 dBi antennas: \[P_r = 0.1 \times \left(\frac{0.333}{4\pi \times 1000}\right)^2 = 7.0 \times 10^{-11} \text{ W} = -71.5 \text{ dBm}\]
This means LoRaWAN (receiver sensitivity -137 dBm) has 65.5 dB link margin at 1 km, while Wi-Fi (sensitivity -90 dBm) has 18.5 dB margin. At 10 km, free-space received power drops to -91.5 dBm – well within LoRaWAN’s reach but 1.5 dB below Wi-Fi’s noise floor. The coverage-bandwidth tradeoff is physics, not marketing.
Quadrant 2 (High Bandwidth, Low Coverage): Local high-speed - Wi-Fi: High data rates, limited range - Ethernet: Maximum bandwidth, wired - Use cases: Video streaming, building automation, industrial equipment
Quadrant 3 (Low Bandwidth, Low Coverage): Personal connectivity - Bluetooth/BLE: Short-range personal devices - Zigbee/Thread: Mesh networking for homes - Use cases: Wearables, home automation, medical devices
Figure 62.2: Bandwidth vs coverage trade-off for different IoT protocols
Figure 62.3: Coverage characteristics of various IoT network technologies
Quadrant 4 (Low Bandwidth, High Coverage): Wide-area sensing - LoRaWAN/Sigfox: Long range, low power, low cost - NB-IoT/LTE-M: Cellular-based LPWAN - Use cases: Agriculture, smart cities, asset tracking, utilities
Concept Check: Bandwidth-Coverage Trade-off
62.4 Network Classification
Time: ~10 min | Difficulty: Intermediate | Reference: P07.C11.U08
Network technologies and protocols can be mapped to traditional network classifications as PAN (Personal Area Network), LAN (Local Area Network), and WAN (Wide Area Network).
Figure 62.4: Network classification by range: PAN, LAN, and WAN technologies
62.4.1 Personal Area Network (PAN)
PAN or Wireless PAN (WPAN) is a network with a small geographical area coverage, for devices such as sensors that require communication within a few meters.
Characteristics:
Range: 1-100 meters
Power: Low to very low (battery-powered)
Data rates: Low to medium (kbps to Mbps)
Topology: Star or mesh
Cost: Very low
Most Popular WPAN Technologies for IoT:
Bluetooth Low Energy (BLE): Wearables, beacons, health devices
Zigbee: Home automation, lighting, security
Z-Wave: Home automation (competing with Zigbee)
Thread: IP-based mesh for smart homes
6LoWPAN: IPv6 over low-power networks
NFC: Contactless payment, access control
Figure 62.5: BLE Personal Area Network topology with smartphone coordinator
62.4.2 Local Area Network (LAN)
LAN provides connectivity within a building or campus, typically covering hundreds of meters to a few kilometers.
Characteristics:
Range: 100 meters - 1 km
Power: Mains powered (or PoE for Ethernet)
Data rates: Medium to very high (Mbps to Gbps)
Topology: Star (for Wi-Fi/Ethernet)
Cost: Low to medium
Most Popular LAN Technologies for IoT:
Wi-Fi (802.11 a/b/g/n/ac/ax): High bandwidth, building coverage
Wi-Fi HaLow (802.11ah): Extended range Wi-Fi for IoT
Figure 62.6: LAN topology with Wi-Fi and Ethernet connected IoT devices
62.4.3 Wide Area Network (WAN)
WAN provides connectivity over large geographical areas, from city-wide to global coverage.
Characteristics:
Range: > 1 km to global
Power: Low (for LPWAN) to medium (for cellular)
Data rates: Very low to very high (bps to Gbps depending on technology)
Topology: Star (for LPWAN) or cellular
Cost: Low (for unlicensed LPWAN) to high (for cellular)
Most Popular WAN Technologies for IoT:
LoRaWAN: Unlicensed, long-range, low-power
Sigfox: Ultra-low bandwidth, very long range
NB-IoT: Cellular LPWAN, licensed spectrum
LTE-M: Cellular with higher bandwidth than NB-IoT
5G: Next-generation cellular for massive IoT
Figure 62.7: Smart city LoRaWAN WAN topology with distributed sensors and gateways
62.5 Terminology Reference
Common IoT Network Terminology
Abbreviation
Full Name
Description
BLE
Bluetooth Low Energy
Low-power Bluetooth for IoT
LAN
Local Area Network
Building/campus network
NFC
Near Field Communication
Very short range (cm)
VSAT
Very Small Aperture Terminal
Satellite communication
BW
Bandwidth
Data carrying capacity
LoWPAN
Low-power Wireless PAN
IPv6 over low-power networks
PAN
Personal Area Network
Short-range network
WAN
Wide Area Network
Long-range network
ISM
Industrial, Scientific, Medical
Unlicensed radio bands
LPWAN
Low Power Wide Area Network
Long range, low power
RFID
Radio Frequency Identification
Tag-based identification
WPAN
Wireless Personal Area Network
Wireless PAN
MAC
Medium Access Control
Layer 2 addressing
PHY
Physical layer
Layer 1 signaling
CSMA/CA
Carrier Sense Multiple Access/Collision Avoidance
Listen-before-talk
QoS
Quality of Service
Performance guarantees
PoE
Power over Ethernet
Power delivery via Ethernet
62.6 Visual Reference Gallery
AI-Generated Figure Variants for Network Access and Physical Layer
These AI-generated SVG figures provide alternative visual representations of network access and physical layer concepts. Each figure uses the IEEE color palette for consistency.
Network Access Physical Layer Overview
Figure 62.8: Network Access Overview: Physical and data link layer components in IoT networks.
Frame Structure
Figure 62.9: Frame Structure: Anatomy of data link layer frames including headers, payload, and error checking.
MAC Address Purpose
Figure 62.10: MAC Address Purpose: How 48-bit MAC addresses enable local network frame delivery.
Visual: Electromagnetic Spectrum for Wireless
Electromagnetic Spectrum
Understanding the electromagnetic spectrum helps explain why different IoT protocols use specific frequency bands and the propagation characteristics that result.
62.7 How It Works: Network Classification in Practice
How It Works: Mapping IoT Devices to Network Classifications
Understanding which network classification (PAN, LAN, or WAN) applies to your IoT device follows a systematic approach:
Step 1: Determine Coverage Range
Measure the physical distance between device and gateway/coordinator
PAN: 1-100 meters (within room or building floor)
LAN: 100m-1km (across building or campus)
WAN: >1km (city-wide or regional)
Step 2: Evaluate Data Rate Requirements
Calculate bytes per transmission × transmissions per second
Match against protocol capabilities (PAN: kbps-Mbps, LAN: Mbps-Gbps, WAN: bps-kbps)
Step 3: Assess Power Constraints
Battery-powered → favor low-power options (BLE in PAN, LoRaWAN in WAN)
Mains-powered → can use higher-power protocols (Wi-Fi in LAN, Cellular in WAN)
Step 4: Consider Deployment Economics
Calculate total cost: infrastructure + subscriptions + maintenance
PAN: Low infrastructure, no subscriptions
LAN: Medium infrastructure, no subscriptions
WAN: Varies (unlicensed LPWAN vs licensed cellular)
Example: A soil moisture sensor needs to transmit 50 bytes every 10 minutes from a field 5 km from the nearest building. Coverage range (5 km) immediately points to WAN. Data rate is tiny (50 bytes / 600 sec = 0.67 bps), favoring LPWAN over cellular. Battery operation for 5+ years rules out cellular (too much power). Conclusion: LoRaWAN (WAN category, unlicensed LPWAN).
Try It: Network Classification Tool
Show code
viewof rangeMeters = Inputs.range([1,50000], {label:"Maximum device range (meters)",step:50,value:500})viewof dataRateBps = Inputs.range([1,10000000], {label:"Required data rate (bps)",step:100,value:1000})viewof batteryPowered = Inputs.select(["Battery-powered","Mains-powered"], {label:"Power source",value:"Battery-powered"}){// Determine network classification from rangelet tier, protocol, rationale;if (rangeMeters <=100) { tier ="PAN (Personal Area Network)";if (batteryPowered ==="Battery-powered") { protocol = dataRateBps >500000?"Bluetooth 5 (BLE)":"Zigbee / BLE"; rationale ="Short range and battery operation: PAN is ideal. BLE and Zigbee offer years of battery life within 100 m."; } else { protocol ="Wi-Fi or Zigbee"; rationale ="Short range and mains power: either PAN (Zigbee) or lightweight LAN (Wi-Fi) works. Wi-Fi enables direct cloud connectivity."; } } elseif (rangeMeters <=1000) { tier ="LAN (Local Area Network)";if (batteryPowered ==="Battery-powered") { protocol ="Wi-Fi HaLow (802.11ah) or Zigbee mesh"; rationale ="Building-scale range with battery: Zigbee mesh can extend to ~300 m via multi-hop; Wi-Fi HaLow reaches ~1 km at lower power than standard Wi-Fi."; } else { protocol = dataRateBps >50000000?"Ethernet (PoE)":"Wi-Fi"; rationale ="Building-scale range and mains power: Wi-Fi for wireless flexibility; Ethernet for guaranteed bandwidth (cameras, industrial equipment)."; } } else { tier ="WAN (Wide Area Network)";if (batteryPowered ==="Battery-powered") { protocol = dataRateBps >100000?"LTE-M (Cat-M1)":"LoRaWAN or NB-IoT"; rationale = dataRateBps >100000?"Wide area + battery + moderate data rate: LTE-M offers up to 375 kbps with low-power sleep modes.":"Wide area + battery + low data rate: LoRaWAN reaches 10+ km with multi-year battery life; NB-IoT works in areas with cellular coverage."; } else { protocol = dataRateBps >1000000?"4G LTE / 5G":"LTE-M or NB-IoT"; rationale = dataRateBps >1000000?"Wide area + mains power + high data rate: 4G/5G cellular provides Mbps throughput with city-wide coverage.":"Wide area + mains power + low data rate: LTE-M or NB-IoT are cost-effective; no battery constraint means power is not the limiting factor."; } }const dataRateFormatted = dataRateBps >=1000000?`${(dataRateBps /1000000).toFixed(2)} Mbps`: dataRateBps >=1000?`${(dataRateBps /1000).toFixed(1)} kbps`:`${dataRateBps} bps`;const rangeFormatted = rangeMeters >=1000?`${(rangeMeters /1000).toFixed(1)} km`:`${rangeMeters} m`;returnhtml` <div style="background:#f0f7f4;border-left:4px solid #16A085;padding:1rem;border-radius:4px;margin-top:0.5rem;"> <div style="font-size:0.85rem;color:#555;margin-bottom:0.5rem;"> Range: <strong>${rangeFormatted}</strong> | Data rate: <strong>${dataRateFormatted}</strong> | Power: <strong>${batteryPowered}</strong> </div> <div style="font-size:1.1rem;font-weight:bold;color:#2C3E50;margin-bottom:0.4rem;"> Classification: ${tier} </div> <div style="color:#16A085;font-weight:600;margin-bottom:0.5rem;"> Recommended Protocol: ${protocol} </div> <div style="color:#555;font-size:0.9rem;">${rationale}</div> </div> `;}
62.8 Worked Example: Designing a Multi-Tier Network for a Smart Hospital
Scenario: A 200-bed hospital needs to connect four categories of IoT devices. Each category has distinct range, data rate, power, and reliability requirements. Your task is to select the appropriate network classification (PAN, LAN, or WAN) and specific technology for each.
300 m range exceeds Zigbee’s 100 m direct range, but Zigbee mesh routing extends coverage through hop-by-hop relay
500 sensors across 12 floors: ~42 sensors per floor. With mesh, every sensor is a potential router.
At 0.4 bps average, the 250 kbps Zigbee channel can support all 500 sensors simultaneously (total load: 200 bps = 0.08% channel utilization)
Battery life with 5-minute interval: Zigbee sleep current 1 uA + wake every 5 min for 10 ms at 20 mA = 5+ years on 2x AA batteries
Security cameras – Ethernet PoE (LAN):
2-5 Mbps per camera. 80 cameras = 160-400 Mbps aggregate.
Gigabit Ethernet switches with PoE+ (IEEE 802.3at): 25.5W per port, sufficient for pan-tilt-zoom cameras
PoE budget: 80 cameras x 15W average = 1,200W. Requires 2x 48-port PoE switches with 740W budget each.
Why not Wi-Fi? Hospital RF environment has interference from medical equipment. Wired provides guaranteed bandwidth and no RF congestion.
Ambulance tracking – LTE-M (WAN):
City-wide mobile coverage requires cellular.
LTE-M supports handover between cell towers (NB-IoT does not handle mobility well).
64 bytes every 30 seconds = 2.1 bps. LTE-M’s 1 Mbps capacity is 475,000x more than needed.
Monthly data: 64 bytes x 2,880 readings/day x 30 days = 5.5 MB/month. At $2/month per SIM = $600/year for fleet.
Step 3: Infrastructure cost estimate
Component
Unit Cost
Qty
Total
BLE patient monitors
$45 each
200
$9,000
BLE nurse station receivers (1 per floor)
$120 each
12
$1,440
Zigbee sensors
$25 each
500
$12,500
Zigbee coordinators (1 per floor)
$85 each
12
$1,020
PoE cameras (1080p, PTZ)
$350 each
80
$28,000
48-port PoE+ switches
$2,200 each
2
$4,400
LTE-M modules + antennas
$20 each
25
$500
LTE-M SIM plans (year 1)
$24/yr each
25
$600
Year 1 Total
$57,460
Key insight: This hospital uses all three network classifications simultaneously – PAN for patient wearables (short-range, low-power), LAN for building infrastructure (medium-range, high-bandwidth), and WAN for mobile fleet (wide-area coverage). The total cost is dominated by the LAN tier (cameras + switches = $32,400, 56% of budget) because high-bandwidth video requires the most expensive infrastructure. The 500 environmental sensors cost less than the 80 cameras despite having 6x more devices – because low-bandwidth sensors need cheap hardware while video cameras need expensive optics and PoE switches.
62.9 Incremental Complexity Examples
Beginner Level: Home Smart Lighting (PAN Only)
Scenario: A homeowner wants to control 10 smart bulbs in a 3-bedroom house using a smartphone app.
Network Design:
Classification: PAN (Zigbee mesh)
Range: 100m² house, max 15m between any two bulbs
Topology: Mesh (each bulb is a router)
Cost: $25/bulb × 10 + $35 hub = $285
Why this works: All devices within PAN range. No WAN needed since smartphone connects via LAN (home Wi-Fi) to hub, which bridges to Zigbee PAN.
Key Takeaway: For single-building automation with low bandwidth needs, a PAN is sufficient.
Intermediate Level: Office Building (PAN + LAN)
Scenario: 5-floor office needs Wi-Fi for employees, IP cameras for security, and environmental sensors for HVAC control.
Network Design:
PAN tier: 50 Zigbee sensors (temperature, occupancy) on 2-wire bus
Why PAN + LAN: Sensors use Zigbee PAN (low power, low bandwidth). Cameras and employee devices use Wi-Fi LAN (high bandwidth). Both networks coexist without interference (Zigbee at 2.4 GHz with frequency agility, cameras at 5 GHz).
Key Takeaway: Combining PAN for sensors and LAN for high-bandwidth devices optimizes cost and power.
Advanced Level: Smart City Deployment (PAN + LAN + WAN)
Scenario: A city deploys 5,000 parking sensors across 10 km², 200 traffic cameras at intersections, and 50 environmental monitoring stations.
Network Design:
PAN tier (station internals):
Each monitoring station: 5 I2C sensors (air quality, noise, weather) connected to microcontroller
Cost: $200/station × 50 = $10,000
LAN tier (traffic cameras):
200 cameras connected via fiber to 20 local switch cabinets
PAN: Within each station, sensors communicate over short distances (I2C, SPI)
LAN: Cameras need high bandwidth (1-5 Mbps), fiber provides that across the city grid
WAN: Parking sensors and stations are distributed over 10 km² - LoRaWAN is the only practical option for battery-powered devices at this scale
Key Takeaway: Large-scale IoT deployments almost always require mixing all three network classifications. Choose each tier based on range, bandwidth, and power requirements.
62.10 Knowledge Check
Quiz 1: Smart City Parking Sensors
Quiz 2: Warehouse Asset Tracking
Quiz 3: Physical Layer Functions
Quiz 4: MAC Methods
62.11 Concept Relationships
Understanding how network classifications relate to each other and to IoT design decisions:
Concept
Related To
Relationship
PAN (Personal Area Network)
LAN, Protocol Selection
PANs connect to LANs via gateways; choice depends on range/power needs
Bandwidth-Coverage Trade-off
Network Classification, Cost
Inversely proportional: higher bandwidth = shorter range OR higher cost
Multi-Tier Architecture
Scalability, Reliability
Combining PAN+LAN+WAN provides flexibility; each tier optimized for specific needs
Protocol Selection
Data Rate, Power Budget, Distance
Technical requirements constrain viable protocols within each classification
LoRaWAN vs Cellular
WAN, Total Cost of Ownership
Unlicensed (no subscription) vs licensed (reliable, expensive); 10-year TCO differs 3-5×
Wi-Fi vs Ethernet
LAN, Deployment Cost
Wireless (flexible, interference-prone) vs wired (reliable, installation-heavy)
62.12 Practice: Match and Sequence
Common Pitfalls
1. Choosing Technology Based on Classification Label Rather Than Requirements
Calling a deployment a “LAN” does not mean Wi-Fi is the right choice if devices are battery-powered and need 5-year battery life. Fix: always derive technology requirements from device constraints and deployment requirements, not from the network classification label.
2. Confusing LPWAN Range With Urban Indoor Range
LoRaWAN achieves 5–15 km in open rural terrain but only 1–2 km in dense urban environments and far less indoors. Fix: use city-specific or environment-specific range models, not the maximum open-field range from the datasheet.
3. Mixing PAN and LAN Devices Without a Gateway
A Bluetooth PAN sensor cannot communicate directly with a Wi-Fi LAN device. A gateway performing Bluetooth-to-Wi-Fi translation is required. Fix: identify all network classification boundaries in a deployment and ensure gateways are planned for each boundary crossing.
🏷️ Label the Diagram
Code Challenge
62.13 Summary
The IoT landscape uses multiple protocols simultaneously - often combining PAN (Zigbee sensors) to LAN (Wi-Fi gateway) to WAN (cellular/LPWAN) to Cloud. Understanding the strengths and limitations of each protocol enables optimal IoT system design.
Key Takeaways
Network Classifications: | Type | Range | Technologies | Use Cases | |——|——-|————–|———–| | PAN | 1-100m | BLE, Zigbee, Thread, Z-Wave | Wearables, smart home, sensors | | LAN | 100m-1km | Wi-Fi, Ethernet, HaLow | Building automation, cameras | | WAN | >1km | LoRaWAN, Cellular, Satellite | Smart cities, agriculture, utilities |
Bandwidth-Coverage Trade-off:
High bandwidth + High coverage = 5G (expensive)
High bandwidth + Low coverage = Wi-Fi, Ethernet
Low bandwidth + Low coverage = Bluetooth, Zigbee
Low bandwidth + High coverage = LoRaWAN, Sigfox, NB-IoT
Protocol Selection Criteria:
No one-size-fits-all: Match protocol to application requirements
Total Cost of Ownership: Include subscription fees, not just hardware
Security: Encryption, authentication at physical layer where possible
Scalability: Can the network grow with your deployment?
Interoperability: Standard protocols vs proprietary solutions
Best Practices:
Combine multiple network types: PAN sensors -> LAN gateway -> WAN cloud
Design for growth - choose protocols that scale
Consider total cost including operations, not just hardware
Plan for technology evolution and migration paths
62.14 Try It Yourself: Network Classification Design Challenge
Exercise: Design a Three-Tier Network for a University Campus
Scenario: Your university wants to deploy IoT across a 50-hectare campus with 20 buildings. Design a network architecture for these requirements:
Devices to support:
5,000 student wearable badges (attendance tracking, door access) - transmit 20 bytes every 30 seconds
200 lecture hall environmental sensors (CO2, temp, humidity) - transmit 50 bytes every 5 minutes
80 security cameras (1080p) - continuous streaming at 3 Mbps each
15 electric shuttle buses (GPS tracking) - transmit 100 bytes every 10 seconds while in motion
Your task:
Assign each device category to PAN, LAN, or WAN
Choose specific protocols for each tier
Calculate total 5-year cost
Justify your decisions
Hint: Start with Range and Mobility
Consider these factors for each device category: - Range: How far are devices from infrastructure? - Mobility: Stationary or moving between buildings? - Data rate: Calculate bytes/second required - Power: Battery or mains? - Economics: Infrastructure cost vs subscription fees over 5 years
Solution: Multi-Tier Campus Network Design
Device 1: Student wearable badges (5,000 units)
Classification: PAN → LAN → WAN hybrid
Design: BLE badges (PAN) connect to 200 BLE-to-Wi-Fi gateways distributed across buildings. Gateways use campus LAN (existing Ethernet) for backhaul.
Rationale: Badges must work indoors (rules out LPWAN). BLE PAN consumes minimal power (18-month battery life). Existing campus LAN avoids WAN subscription costs.
