11 IoT Glossary: Terms From W Through Z
11.1 Start With the Decision
WebSockets keep one TCP link open for traffic in both directions. Wi-Fi moves local data by radio.
11.2 Route Overview
This is part 2 of 2. Review IoT Glossary: Terms From S Through V for the preceding evidence.
11.3 Learning Objectives
- Test w with a concrete scenario and pass criteria.
- Validate match the wireless protocol to its key trait with a concrete scenario and pass criteria.
11.4 Chapter Roadmap
- W
- Z
- Additional Reference Terms
- Common Pitfalls
- 1. Confusing QoS 1 ‘at least once’ with ‘exactly once’
- 2. Using Zigbee and Wi-Fi simultaneously without interference management
- 3. Choosing Wi-Fi for every sensor node
- Label the Diagram
- Code Challenge
- Worked Example: Selecting QoS Levels for a Multi-Sensor IoT System
- Decision Framework: Wireless Protocol Selection Using Glossary Terms
- Common Mistake: Misunderstanding Time Series Database Requirements
- Concept Relationships
- How These Terms Interconnect
- Match the Wireless Protocol to Its Key Trait
- Order the TLS Handshake Steps
- Summary
- See Also
- Related Resources
- What’s Next
- Navigation
- Key Takeaway
11.5 W
11.5.1 WebSocket
Definition: A protocol providing full-duplex communication channels over a single TCP connection, enabling real-time bidirectional data exchange between web clients and servers.
In simple terms: A two-way communication line - unlike HTTP polling, WebSocket keeps a connection open so servers can push updates as events happen.
Common confusions:
- WebSocket vs. HTTP - WebSocket is persistent and bidirectional, HTTP is request-response
- WebSocket starts with HTTP upgrade, then switches protocol
Related terms: HTTP, MQTT, Real-time
Synonyms: WS, WSS (secure WebSocket)
11.5.2 Wi-Fi
Definition: A family of wireless networking protocols based on IEEE 802.11 standards, providing local area network connectivity using radio frequencies in 2.4 GHz and 5 GHz bands.
In simple terms: Wireless internet - the most common way to connect devices to a network without cables, using radio waves to link devices to a router or access point.
Common confusions:
- Wi-Fi vs. internet - Wi-Fi is the local connection, internet is what you access through it
- Wi-Fi generations (4/5/6/6E/7) have different speeds and capabilities
Related terms: BLE, Zigbee, Bandwidth
Synonyms: Wi-Fi, 802.11, WLAN
11.5.3 Wildcard
Definition: A special character in topic patterns that matches multiple topic levels, enabling subscribers to receive messages from multiple topics with a single subscription.
In simple terms: A pattern matcher - like using ”*” to find all files in a folder, wildcards let you subscribe to groups of related topics at once.
Common confusions:
- MQTT uses + (single level) and # (multi-level) wildcards
- Different protocols have different wildcard syntax
Related terms: Topic, Subscription, Filter
Synonyms: topic wildcard, pattern matching
11.6 Z
11.6.1 Zigbee
Definition: A wireless mesh networking protocol based on IEEE 802.15.4 for low-power, low-data-rate applications, widely used in home automation and industrial control.
In simple terms: A mesh network for smart homes - devices form a web of connections where messages hop from device to device, creating reliable coverage throughout your home.
Common confusions:
- Zigbee vs. Z-Wave - different protocols, incompatible, both used for home automation
- Zigbee vs. Thread - Zigbee has its own network layer, Thread uses IPv6
Related terms: Z-Wave, Mesh, Thread, Matter
Synonyms: Zigbee, IEEE 802.15.4
11.6.2 Z-Wave
Definition: A proprietary wireless mesh networking protocol operating in sub-GHz frequencies, designed for smart home automation with low interference and excellent building penetration.
In simple terms: Another smart home mesh - similar purpose to Zigbee but uses different radio frequencies that pass through walls better, with strictly controlled certification.
Common confusions:
- Z-Wave vs. Zigbee - incompatible protocols, Z-Wave uses sub-GHz, Zigbee uses 2.4 GHz
- Z-Wave is proprietary with mandatory certification, Zigbee is more open
Related terms: Zigbee, Mesh, Matter
Synonyms: none common
11.6.3 6LoWPAN
Definition: IPv6 over Low-Power Wireless Personal Area Networks, an adaptation layer enabling IPv6 packets to be transmitted over IEEE 802.15.4 networks through header compression and fragmentation.
In simple terms: Shrinking IPv6 for tiny devices - makes full internet protocol (IPv6) work on constrained networks by compressing headers and splitting large packets.
Common confusions:
- 6LoWPAN is an adaptation layer, not a complete protocol stack
- Used by Thread and often paired with CoAP
Related terms: CoAP, Thread, MTU
Synonyms: IPv6 over LoWPAN
11.7 Additional Reference Terms
11.7.1 Cellular IoT
Definition: IoT connectivity using cellular networks (LTE-M, NB-IoT, 5G), leveraging existing mobile infrastructure for wide-area coverage and managed connectivity services.
In simple terms: Using cell phone networks for IoT - connecting devices through the same towers your phone uses, providing coverage almost anywhere without deploying your own network.
Common confusions:
- Cellular IoT includes multiple technologies (NB-IoT, LTE-M, 5G mMTC)
- Requires carrier subscription and SIM/eSIM
Related terms: NB-IoT, LTE-M, LPWAN
Synonyms: mobile IoT, cellular M2M
11.7.2 IIoT (Industrial IoT)
Definition: Application of IoT technologies in industrial settings for manufacturing, energy, transportation, and other industrial processes, emphasizing reliability, safety, and interoperability.
In simple terms: IoT for factories and industry - connecting machines, sensors, and systems in industrial environments to improve efficiency, safety, and maintenance.
Common confusions:
- IIoT vs. Industry 4.0 - IIoT is the connectivity, Industry 4.0 is the broader transformation
- IIoT has stricter requirements for reliability and safety than consumer IoT
Related terms: OPC-UA, Modbus, Digital Twin
Synonyms: Industrial Internet of Things, Industry 4.0 IoT
11.7.3 Real-time
Definition: System responses occurring within defined time constraints, where data is processed and delivered within the deadline needed for action or feedback.
In simple terms: Instant response - systems that react immediately, like brake lights that come on the moment you press the pedal, not seconds later.
Common confusions:
- Real-time doesn’t mean “fast” - it means “predictably within time constraints”
- Hard real-time vs. soft real-time have different tolerance for delays
Related terms: Latency, DDS, WebSocket
Synonyms: real-time systems, low-latency
11.7.4 Frequency
Definition: The rate at which a periodic signal oscillates, measured in Hertz (Hz), determining radio band allocation for wireless communication and signal characteristics.
In simple terms: How fast waves wiggle - measured in cycles per second, different frequencies are used for different purposes (2.4 GHz Wi-Fi, 915 MHz LoRa, etc.).
Common confusions:
- Higher frequency generally means more bandwidth but less range/penetration
- Different regions use different frequency bands (regulatory)
Related terms: Channel, Bandwidth, Wi-Fi
Synonyms: radio frequency, Hz
Common Pitfalls
These pitfalls connect glossary terms to operational consequences. Read them as three review paths: delivery semantics must be completed by application idempotency, shared spectrum needs measured channel planning, and radio choice must follow the workload’s range and power budget. In each case, the familiar label is only a starting point; the deployment evidence decides whether the design is safe.
MQTT QoS 1 asks the broker and client to complete at-least-once delivery, so duplicate messages are still possible. If a sensor reading is published with QoS 1 and the broker acknowledgment is lost, the device can retransmit, causing a subscriber to receive the same reading twice. For actuator commands where duplicates could cause repeated actions, use command ids, idempotent state transitions, acknowledgement state, and application-level safety checks; MQTT QoS 2 only narrows duplicate delivery within the MQTT exchange.
Zigbee commonly uses 2.4 GHz IEEE 802.15.4 channels 11-26, while Wi-Fi channels in the same band occupy wider 20/22 MHz slices. Poor channel planning can reduce packet delivery, especially when a Zigbee channel sits inside a busy Wi-Fi channel. Survey the site, separate busy Wi-Fi and Zigbee channels where possible, check device support for upper Zigbee channels such as 25 or 26, and consider sub-GHz options such as Z-Wave only after checking regional frequency and certification constraints.
Wi-Fi offers high throughput, but it is often a poor fit for small battery-powered sensors that send short messages. For low-rate sensing, compare Wi-Fi with Zigbee, Thread, Z-Wave, Sigfox, LoRaWAN, and Cellular IoT before choosing. Match the wireless technology to range, power budget, topology, and infrastructure constraints instead of defaulting to the most familiar network.
You’re designing an industrial monitoring system with 50 sensors transmitting via MQTT. The numbers below are worksheet assumptions for comparing choices; replace them with measured radio current, broker logs, and failure costs in a real design.
System Overview:
Begin with the traffic classes: 30 temperature sensors report at 1-minute intervals, 15 vibration sensors report equipment-health data every 10 seconds, and 5 emergency-stop buttons send immediate alerts. Their different consequences and repetition rates are why one QoS setting cannot be assumed for the whole system.
Step 1: Temperature Telemetry
Each temperature sensor produces 1,440 readings per day, or 43,200 readings across 30 sensors. QoS level 0 provides at-most-once, fire-and-forget delivery. Losing one isolated reading from this high-frequency trend may be acceptable only when aggregation, alerting, and data-quality rules tolerate gaps. On those worksheet assumptions, choose QoS 0: one MQTT transmission per reading and no acknowledgement exchange, which reduces radio airtime relative to QoS 1 or QoS 2.
Step 2: Vibration Monitoring
Vibration monitoring is predictive-maintenance evidence, so a missed high-vibration event could delay intervention; quantify that consequence from the asset and workflow. QoS level 1 provides acknowledged at-least-once delivery, but duplicates remain possible. Tag each reading with a timestamp and sequence number, then use idempotency checks at the backend. The resulting decision is QoS 1 with deduplication: roughly twice the MQTT message legs of QoS 0 in this simplified comparison, justified only by the measured maintenance risk.
Step 3: Emergency Stop Buttons
Life-safety commands need controlled execution, explicit acknowledgement, and a safe final state. QoS level 2 narrows duplicate delivery through the MQTT exactly-once exchange, but duplicate and missed-stop handling still belong in the application safety design. Follow the protocol in order: PUBLISH travels from client to broker; PUBREC confirms receipt; PUBREL releases the exchange; and PUBCOMP marks it complete. Choose QoS 2 plus command idempotency and safety interlocks only where this extra handshake is justified and the full end-to-end stop path still meets its timing requirement.
Step 4: Gateway Heartbeats
Gateway heartbeats are telemetry for online or offline health, sent every 60 seconds. If the stale-state policy accepts that detection interval, a lost heartbeat is replaced by the next scheduled observation. The bounded decision is therefore QoS 0 for redundant, fire-and-forget pings, with the 60-second delay made visible to operators.
Resulting MQTT Configuration Table
| Message Type | QoS | Rationale | Overhead | Energy Note |
|---|---|---|---|---|
| Temperature telemetry | 0 | Redundant trend data | 1x | Lowest acknowledgement traffic |
| Vibration data | 1 | Critical alerts, deduplication via timestamp and sequence | 2x | Higher radio airtime, justified by maintenance risk |
| E-stop commands | 2 + app safety | Critical command path | 4x MQTT handshake | Rare traffic; verify full safety timing |
| Gateway heartbeats | 0 | Status pings, 60s redundancy | 1x | Missing one heartbeat can be handled by stale-state rules |
Relative Airtime Impact
If we incorrectly used QoS 2 for temperature telemetry:
At QoS 0, the worksheet’s 43,200 daily temperature messages require 43,200 MQTT transmission legs. Treating the four-step QoS 2 exchange as four legs would produce 172,800 legs, four times the simplified count. That does not by itself prove four times the battery use: it predicts more airtime, broker work, battery drain, and congestion exposure, which must be converted into battery-life and maintenance cost using measured device current.
Key Lesson: Match QoS level to message criticality. The glossary’s “Common confusions” section for QoS emphasizes: “Higher QoS means more overhead and latency” - reserve QoS 2 only for irreplaceable commands.
11.7.5 Interactive QoS Impact Calculator
Experiment with different QoS levels to see their impact on battery life and network overhead:
Use this framework to compare Wi-Fi, Zigbee, Thread, BLE, LoRaWAN, and Cellular IoT. Treat the ranges and lifetimes as starting estimates to verify with the exact radio, antenna, enclosure, firmware duty cycle, region, and site survey.
11.7.6 Decision Criteria Table
| Requirement | If This Applies | Consider | Glossary Terms to Review |
|---|---|---|---|
| Short-range personal area | Wearables, room sensors | BLE | BLE, Frequency |
| Room/building automation | Home/office automation | Zigbee, Thread, Wi-Fi | Mesh, Star Topology |
| Campus/farm/site coverage | Sparse telemetry with gateway planning | LoRaWAN, Sigfox where service exists | LPWAN |
| Carrier-managed wide-area coverage | City-wide, regional, mobile assets | NB-IoT, LTE-M | Cellular IoT |
| Multi-year battery target | Infrequent reporting | LoRaWAN, Sigfox, NB-IoT after current-budget validation | LPWAN, Duty Cycle |
| Throughput > 1 Mbps | Video, audio streaming | Wi-Fi | Bandwidth, Throughput |
| Latency < 100 ms | Real-time control | Wi-Fi, Thread | Latency, Real-time |
| Mesh networking required | Redundant paths and self-healing routes | Zigbee, Thread | Mesh, Routing |
| IP connectivity required | Directly addressable devices | Wi-Fi, Thread, 6LoWPAN | 6LoWPAN, Gateway |
| No infrastructure | Deploy in remote areas | LoRaWAN (own gateway), avoid NB-IoT (needs carrier) | LPWAN |
11.7.7 Comparative Analysis Example
Scenario: Smart building with 200 occupancy sensors, 50-meter coverage, 10-year deployment
| Protocol | Range | Battery Life | Data Rate | Mesh? | Verdict |
|---|---|---|---|---|---|
| Wi-Fi | 50m | 6 months | 54 Mbps | No | Battery too short |
| BLE | 10m | 2 years | 1 Mbps | No | Range insufficient |
| Zigbee | 100m | 5 years | 250 kbps | Yes | Best fit |
| Thread | 100m | 5 years | 250 kbps | Yes | Alternative with IPv6 benefit |
| LoRaWAN | 5 km | 10 years | 50 kbps | No | Overkill for 50m, higher cost |
Decision: Zigbee or Thread
Zigbee: Mature ecosystem, existing building automation standards (ZHA, ZCL) Thread: Future-proof with IPv6, compatible with Matter Recommendation: Thread if using Matter smart home integration, Zigbee otherwise
11.7.8 Common Confusion Resolution
Question: “Should I use Zigbee or Z-Wave?”
From glossary “Common confusions” sections:
Zigbee: 2.4 GHz profiles are common, with mesh routing and larger theoretical address space than most residential deployments need Z-Wave: sub-GHz regional frequencies, managed interoperability certification, and controller/device limits that depend on the product generation
Decision factors:
Step 1 — Building penetration: Z-Wave (868/915 MHz) penetrates walls better than Zigbee (2.4 GHz) Step 2 — Wi-Fi interference: Z-Wave avoids 2.4 GHz congestion Step 3 — Scale: Zigbee for >100 devices, Z-Wave for <100 Step 4 — Global deployment: Zigbee (single frequency), Z-Wave requires region-specific modules
Rule of thumb: For residential deployments with sub-GHz region support, Wi-Fi interference, and strong wall-penetration needs, evaluate Z-Wave. For larger commercial deployments or global product planning, evaluate Zigbee or Thread with explicit channel planning, controller support, and certification checks.
The Problem: Students often say “I’ll use InfluxDB” or “I’ll use TimescaleDB” without understanding when a TSDB is actually needed.
Teaching Scenario:
A team stored 3 temperature sensors (1-minute sampling) in a MySQL database using:
CREATE TABLE readings (
sensor_id INT,
value FLOAT,
timestamp DATETIME,
PRIMARY KEY (sensor_id, timestamp)
);
After 6 months, queries like “average temperature per hour for the last week” took 45 seconds. With 12 million rows, the database consumed 2.1 GB.
Why This Can Fail:
Step 1 — Wrong query shape for the index: A simple timestamp table can work, but long time-range aggregation may need partitioning, rollups, or a purpose-built store. Step 2 — No downsampling policy: Raw data kept forever can make old queries expensive when older data only needs hourly or daily summaries. Step 3 — No compression or retention plan: TSDBs often provide compression and retention controls that general tables need to be configured to match.
How to Decide: Do You Need a TSDB?
Use this decision guide:
| Condition | Recommended storage choice |
|---|---|
| Data is not time-stamped | Use PostgreSQL, MongoDB, or another general-purpose database. |
| Time-stamped data below 10,000 readings/day | Use PostgreSQL with a timestamp index if query needs are simple. |
| Time-stamped data from 10,000 to 1M readings/day | Consider a TSDB, such as TimescaleDB for SQL familiarity or InfluxDB for IoT-native workflows. |
| Time-stamped data above 1M readings/day | Use a TSDB such as InfluxDB, Prometheus, QuestDB, or TimescaleDB. |
Corrected Design Using TSDB:
Option 1: TimescaleDB (if team knows SQL)
Glossary: TimescaleDB = PostgreSQL extension for Time Series Advantages: Familiar SQL syntax, use existing PostgreSQL knowledge Automatic chunking: Partitions data by time (1-week chunks) Continuous aggregates: Pre-compute hourly/daily averages in background Compression: 10x reduction using columnar storage plus delta encoding
Option 2: InfluxDB (purpose-built for IoT)
Glossary: InfluxDB = Native TSDB with custom query language (Flux) Advantages: Designed for high-throughput Telemetry ingestion workflows and built-in downsampling Tagging system: Store sensor metadata (location, type) as tags, not separate tables Retention policies: Auto-delete old data (keep 7 days raw, 90 days hourly, 2 years daily)
Illustrative Performance Comparison:
| Database | 6-month Dataset Size | “Last week hourly average” Query Time |
|---|---|---|
| MySQL (original) | 2.1 GB | 45 seconds |
| TimescaleDB (compressed) | 184 MB | 1.2 seconds |
| InfluxDB (with downsampling) | 94 MB | 0.3 seconds |
Treat these values as a teaching comparison for one workload shape, not vendor benchmarks. Real results depend on schema, indexes, hardware, retention policy, compression settings, and query design.
When NOT to Use a TSDB:
Event logs: Use Elasticsearch (full-text search), not InfluxDB Transactional data: Use PostgreSQL for ACID transaction semantics, not TimescaleDB Small datasets: Below 100,000 points total, PostgreSQL with a timestamp index is simpler. Complex joins: If queries need to join with user tables or product catalogs, PostgreSQL may be the better primary store.
Key Lesson from Glossary: Review TSDB “Common confusions”: “TSDB vs. relational database - TSDBs optimize for time-based writes and queries.” Match database to query pattern, not just data type.
11.8 Concept Relationships
Quality of Service Ecosystem:
QoS connects to at-least-once, at-most-once, exactly-once, idempotency, and retry logic. This chain shows how delivery semantics require specific design patterns
Wireless Protocol Hierarchy:
Physical layer: Wi-Fi, BLE, LoRa (radio modulation) Network layer: Zigbee, Thread, 6LoWPAN (routing, addressing) Application layer: MQTT, CoAP, HTTP (data exchange patterns)
Database Decision Tree:
High-rate time series data often requires a TSDB. TSDB options: InfluxDB (purpose-built) vs TimescaleDB (SQL-based) Schema determines query patterns and storage efficiency
Security Stack:
TLS (transport encryption) + Token (authentication) + ACL (authorization) = Secure IoT
11.9 Summary
This section covered 36 essential IoT terms from Q through Z, including four additional reference terms:
| Category | Terms | Examples |
|---|---|---|
| Reliability and messaging | 6 | QoS, Retry, Subscriber, Subscription, Topic, Wildcard |
| Data and databases | 7 | Sampling, Schema, Serialization, Time Series, TimescaleDB, Timestamp, TSDB |
| Security and access | 2 | TLS, Token |
| Hardware and interfaces | 4 | Sensor, SoC, SPI, UART |
| Networking and performance | 6 | REST, Routing, Star Topology, Throughput, WebSocket, Real-time |
| Wireless connectivity | 7 | Sigfox, Thread, Wi-Fi, Zigbee, Z-Wave, 6LoWPAN, Cellular IoT |
| Architecture and reference | 4 | Serverless, Telemetry, IIoT, Frequency |
11.10 See Also
Within This Module:
Glossary A-F - Navigate to the start of the glossary sequence for related terms Glossary A-F - Foundation terms (Authentication, CoAP, Edge Computing) Glossary G-P - Networking and hardware terms (MQTT, Gateway, LoRaWAN)
Deep Dives (Cross-Module):
MQTT QoS Levels - Practical implementation of QoS 0, 1, 2 Zigbee vs Thread Comparison - When to use each mesh protocol Time Series Database Selection - InfluxDB vs TimescaleDB vs QuestDB benchmarks
Hands-On Practice:
TSDB Query Lab - Practice InfluxDB and TimescaleDB queries Wi-Fi vs Zigbee Range Calculator - Compare wireless protocol coverage
11.11 What’s Next
| If you want to… | Read this |
|---|---|
| Look up IoT terms from A through F | Glossary A-F |
| Look up IoT terms from G through P | Glossary G-P |
| Review the glossary from the first range | Glossary A-F |
| Apply security concepts (TLS, QoS) in capstone projects | Capstone Projects |
| Deep dive into time-series database concepts | Time-Series Fundamentals |
11.13 Key Takeaway
Later glossary terms often capture operational concerns such as reliability, telemetry, updates, and wireless behavior. These concepts should appear in test plans and deployment reviews, not only in definitions.
11.14 Continue Your Route
This final part closes the route from W through Key Takeaway. Return to IoT Glossary: Terms From S Through V or continue from the capstone module index.
