Skip to content

Appendix

A. Glossary

TermEnglish / Full NameDescription
AIoTArtificial Intelligence of ThingsDeep fusion of AI and IoT — from passive connectivity to active intelligence
MCPModel Context ProtocolOpen standard (Anthropic, 2024) for interaction between AI and external tools and data sources; donated in December 2025 to the Agentic AI Foundation under the Linux Foundation
Tool callingTool callingMechanism by which an LLM invokes external tools (e.g., devices) via function calls
RAGRetrieval-Augmented GenerationModel answers augmented with retrieved knowledge
AgentAI AgentIntelligent agent that perceives, reasons, plans, and executes multi-step tasks
Thing modelThing Model / ProfileAbstraction of device capabilities (properties/services/events) that hides protocol differences
Point valuePoint ValueSemantically tagged device data point (device ID + timestamp + unit + value)
Agentic CenterAgentic CenterIoT DC3's intelligent decision hub, built on Spring AI
Bounded autonomyBounded AutonomyAn agent executes multi-step tasks autonomously within explicit boundaries of permissions, policies, confirmations, and audits; the clearer the boundaries, the more authority can be delegated — safety-critical decisions always remain with humans
LPWANLow-Power Wide-Area NetworkLow-power wide-area network technologies such as NB-IoT and LoRa
RedCapReduced CapabilityLightweight 5G (Rel-17) for mid-tier IoT
TSFMTime Series Foundation ModelTime-series foundation models (e.g., TimesFM, Chronos) for zero-shot forecasting
On-device SLMSmall Language Model (SLM)A small language model below a few billion parameters that, once quantized, can be pushed down to edge gateways to support lightweight semantic tasks such as device Q&A, alarm summarization, and first-pass work-order screening
Device shadowDevice ShadowDesired/actual device state maintained by the platform, decoupled from online status
OTAOver-the-Air UpdateRemote over-the-air update of firmware/software; must be paired with signature verification, encrypted transport, and anti-rollback protection — otherwise a single malicious update can compromise devices in batches
Edge-cloud collaborationEdge-Cloud CollaborationTiered collaboration: real-time processing at the edge, deep compute in the cloud
RBAC/ABACRole/Attribute-Based Access ControlAccess control based on roles or attributes
MQTTMessage Queuing Telemetry TransportDe facto standard messaging protocol for the IoT
CoAPConstrained Application ProtocolLean web protocol for constrained devices (RFC 7252)
LwM2MLightweight M2MLightweight device-management protocol defined by OMA (over CoAP)
OPC UAOPC Unified ArchitectureIndustrial interoperability application-layer protocol (IEC 62541)
QoSQuality of ServiceMessage-delivery semantics: at most once (0) / at least once (1) / exactly once (2)
Time-series databaseTime Series DatabaseStorage and aggregation tailored to timestamped data (TimescaleDB, InfluxDB, etc.)
DIDDecentralized IdentifierDecentralized identifier (W3C standard) controlled by the subject itself
Verifiable credential (VC)Verifiable Credential (VC)Tamper-proof digital credential standardized by W3C: signed by the issuer, held by the holder, verified by the verifier; often combined with DID for device and subject identity. Data Model 2.0 became a W3C Recommendation in May 2025
Federated learningFederated LearningMechanism for jointly training a model across parties without uploading raw data
Consortium / permissioned chainConsortium Chain / Permissioned ChainA blockchain in which only authorized nodes take part in consensus and read/write; the form jointly governed by several known organizations is called a consortium chain (e.g., Hyperledger Fabric), suited to cross-organization collaboration
Smart contractSmart ContractA program deployed on a blockchain that executes automatically once its conditions are met; in IoT, mostly used for on-chain evidence retention and automated trusted execution
Merkle treeMerkle TreeA tree structure that aggregates the hashes of a batch of data pairwise, layer by layer, converging to a root hash; verifying a single record takes only a logarithmic path of hashes — well suited to bandwidth-constrained IoT evidence retention
OracleOracleA bridge service that delivers off-chain data and events to on-chain smart contracts in a trustworthy manner; its own trustworthiness and degree of decentralization are the key risk points in on-chain decision paths
Zero-knowledge proof (ZKP)Zero-Knowledge Proof (ZKP)A cryptographic technique in which a prover convinces a verifier that an assertion is true while revealing nothing beyond the assertion itself; suited to compliance checks of the form "verify the condition holds without exposing the value"
Differential privacy (DP)Differential Privacy (DP)Injects quantifiable noise into query results or training gradients so that whether any single record participates cannot be inferred; suited to group statistics, not to single-point control
Secure multi-party computation (MPC)Secure Multi-Party Computation (MPC)Several parties jointly compute an agreed function without revealing their own inputs; communication and compute overhead are high, so it is mostly used for low-frequency, high-value joint computation
Trusted execution environment (TEE)Trusted Execution Environment (TEE)A hardware-isolated execution region inside the CPU whose code and data are shielded from the host system and even physical probing; performance is close to native, at the cost of having to trust the chip vendor
Homomorphic encryption (HE)Homomorphic Encryption (HE)An encryption scheme that supports computing directly on ciphertext, with the decrypted result matching computation on plaintext; fully homomorphic schemes remain expensive, so current use focuses on specific operators such as ciphertext aggregation
Digital twinDigital TwinReal-time mirror and simulation of a physical entity in digital space
ISA-95ISA-95International standard layered model (L0–L4) for enterprise–control system integration
V2XVehicle-to-EverythingUmbrella term for vehicle-to-vehicle/road/network/person communication (C-V2X/DSRC)
RSU/OBURoadside Unit / On-Board UnitRoadside communication unit / in-vehicle communication unit
Edge computingEdge ComputingPerforming computation and decision-making close to the data source
ADRArchitecture Decision RecordA lightweight document recording the context, options, decision, and consequences of an architecture decision, reviewed and evolved together with code changes
MoSCoWMust / Should / Could / Won't-haveA method that ranks requirements into four tiers — must have, should have, could have, and won't have this time — suited to IoT projects with constrained resources and a clear delivery cadence

B. References

  1. 3GPP TS 22.261 — 5G system service requirements (including IoT scenarios)
  2. 3GPP TS 36.300 — LTE/4G system architecture (overall E-UTRA description)
  3. 3GPP TS 38.300 — Overall description of the 5G NR system architecture
  4. 3GPP TR 38.875 (RedCap study report) / TR 38.821 (NTN non-terrestrial network study report)
  5. CSA, Matter Specification — Unified application-layer standard for smart homes
  6. OASIS, MQTT Version 5.0 — Message Queuing Telemetry Transport
  7. IETF, RFC 7252 (CoAP) / draft-ietf-oauth-v2-1 (OAuth 2.1 draft, not yet finalized) / RFC 8628 (device authorization flow)
  8. Anthropic / Agentic AI Foundation, Model Context Protocol (MCP) Specification — Open standard for AI–tool interaction
  9. Spring reference documentation — Spring Boot 4.0 (GA 2025-11) / Spring Cloud 2025.1 (GA 2025-11) / Spring AI 2.0 (GA 2026-06)
  10. IoT Analytics, State of IoT — Global installed base of IoT devices and industry data
  11. LoRa Alliance, LoRaWAN L2 1.0.4 / 1.1 and Regional Parameters RP-002-1.0.5 (2025-10)
  12. IoT DC3 open-source project — https://gitee.com/pnoker/iot-dc3 (running case throughout the book)
  13. European Union, Regulation (EU) 2024/2847 (Cyber Resilience Act, CRA) — Cybersecurity regulation for products with digital elements (in force 2024-12, phased application 2026-09 / 2027-12)
  14. W3C, Verifiable Credentials Data Model v2.0 (2025-05 W3C Recommendation) and DID Core 1.0 — Standards for decentralized identity and verifiable credentials

C. Index

Protocols & communication: MQTT / CoAP / LwM2M / Modbus / OPC UA / NB-IoT / LoRa(WAN) / 5G (RedCap/NTN) / Wi-Fi / BLE / Zigbee / Matter / Thread / gRPC / REST

Architecture & platform: Five-layer architecture / intelligence layer / microservices / Gateway / Auth / Manager / Data / Agentic / thing model / point value / device shadow / time-series database / message queue / rule engine / edge-cloud collaboration

AI & agents: Large language model (LLM) / Agent / RAG / Tool calling / MCP / Spring AI / Agentic Center / natural-language operations / anomaly detection / predictive maintenance / TSFM / on-device SLM

Security: OAuth 2.1 (IETF draft) / JWT / X.509 certificates / TLS 1.3 / DTLS / RBAC / ABAC / multi-tenancy / prompt injection / PQC (post-quantum cryptography)

Application scenarios: Industrial IoT (IIoT) / digital twins / smart cities / connected vehicles (V2X) / precision agriculture / blockchain + IoT / supply-chain traceability

From Industrial Software to AI Agents · Building a multi-protocol, cloud-native, open-source industrial IoT platform ready to evolve toward AI agents