Skip to content
CampusPrep
ME-702 (B) · Internet of Things/Quick Revision Short Notes

Internet of Things (ME-702 (B)) - Unit 3 Short Notes

UNIT 3: Internet of Things – Comprehensive Short Notes

I. IoT Foundations & Architectural Frameworks

IoT Definition, Ecosystem & Evolution

  • Definition: IoT is a system of interrelated computing devices, mechanical/digital machines, objects, animals, or people provided with unique identifiers (UIDs) and the ability to transfer data over a network without requiring human-to-human or human-to-computer interaction.

    • Core Components:

      • Things: Physical objects with sensors/actuators.

      • Internet: Connectivity infrastructure (IP-based).

      • Analytics: Data processing for insights/decisions.

  • IoT Ecosystem:

    | Layer | Components | Function | |-----------------|-------------------------------------------------|-------------------------------------------| | Device | Sensors, Actuators, Microcontrollers | Data acquisition & action | | Gateway | Edge gateway, Protocol translator | Aggregation, Pre-processing, Security | | Network | LAN/WAN, Cellular, LPWAN | Data transport | | Cloud | Storage, Compute, Databases | Big data analytics, Management | | Application | Dashboards, Mobile/Web apps | Visualization, Control |

  • Evolution M2M → IoT:

    • M2M: Point-to-point, siloed, limited to device-to-device communication (e.g., SCADA).

    • IoT: IP-based, cloud-integrated, scalable, semantic interoperability, data-driven services.

    • Paradigm Shift: From isolated automation to interconnected, intelligent systems with analytics at the core.

  • IoT Gateway Functionality:

    • Protocol Translation: e.g., ZigBee/Modbus ↔ MQTT/HTTP.

    • Edge Processing: Filtering, aggregation, local analytics.

    • Security: Firewall, encryption, authentication at network edge.

    • Device Management: Firmware updates, monitoring.

[!TIP] Exam Focus: Differentiate M2M vs IoT clearly. Gateway functions are high-frequency—list 4 key roles.

Architectural Models & Frameworks

  • IoT Reference Architecture (Common 3/5/7-layer models):

    • Perception Layer: Sensors/actuators (physical world interface).

    • Network Layer: Connectivity (Wi-Fi, BLE, LoRaWAN).

    • Middleware/Service Layer: Data management, service discovery.

    • Application Layer: Domain-specific apps (smart home, health).

    • Business Layer: Business models, policies.

    • Information Model: Standardized data representation (e.g., oneM2M).

  • Service-Oriented Architecture (SOA) for IoT:

    • Services: Reusable, loosely-coupled functional units (e.g., "temperature reading service").

    • Challenges: Resource constraints (memory/CPU), service discovery in dynamic networks, QoS guarantees.

  • Level-Based IoT Systems:

    | Level | Description | Example | |-----------|------------------------------------------|---------------------------------| | Level 3 | Device-centric, local processing, no cloud | Standalone sensor with local alarm | | Level 4 | Cloud-connected, analytics, remote control | Smart thermostat with mobile app |

  • Logical vs Physical Design:

    • Logical: Functional blocks & data flow (e.g., "sensing module → processing module → communication module").

    • Physical: Hardware placement, wiring, deployment topology (e.g., Raspberry Pi + DHT22 sensor + Wi-Fi router).

  • IoT Architectural Framework Characteristics:

    • Scalability, Interoperability, Modularity, Security-by-Design, Support for heterogeneity.

[!TIP] Common Pitfall: Confusing logical design (what it does) with physical design (how it's built). Use examples to clarify.

IoT Characteristics, Challenges & Requirements

  • Core Characteristics:

    • Connectivity: Always-on, IP-based.

    • Heterogeneity: Diverse devices, protocols, platforms.

    • Scalability: Millions of devices.

    • Dynamic & Self-Adapting: Devices join/leave, context-aware.

    • Intelligence: Embedded analytics, ML at edge.

  • Challenges:

    • IoT LAN Issues: Interference (2.4 GHz), limited range, power constraints, security in local networks.

    • Development: Lack of standards, legacy system integration, data silos.

  • Device-Level Requirements:

    • Power: Battery life (years), energy harvesting.

    • Size/Form Factor: Miniaturization.

    • Cost: <$5 for mass deployment.

    • Reliability: 99.9% uptime in harsh environments.

    • Security: Secure boot, hardware encryption.

IoT Enablers & Interdependencies

  • IoT Planes:

    • Device Plane: Sensors/actuators.

    • Communication Plane: Networks/protocols.

    • Service Plane: Data processing, APIs.

    • Management Plane: Device mgmt., security.

    • Application Plane: User-facing apps.

    • Business Plane: Monetization, policies.

  • Enablers:

    • Hardware: MCUs (ESP32), SBCs (Raspberry Pi), sensors.

    • Software: OS (FreeRTOS), middleware.

    • Protocols: MQTT, CoAP, ZigBee.

    • Platforms: AWS IoT, Azure IoT.

    • Analytics: Stream processing (Apache Kafka), ML frameworks.

  • Interdependencies:

    Hardware choice → affects protocol support → influences cloud platform → determines analytics capability.

    Example: Low-power sensor (hardware) → uses ZigBee (protocol) → requires gateway (management) → sends data to AWS IoT (platform) → analyzed by Lambda (analytics).

[!TIP] Diagram Expected: "IoT Planes & Enablers Interdependencies" block diagram showing layers with arrows indicating data/control flow.

II. Hardware Components & Device Design

Sensors

  • Types:

    • Physical: Temperature (thermocouple), Humidity (capacitive), Pressure (piezoresistive), Motion (PIR), Light (LDR), Gas (MQ-series).

    • Chemical/Biological: pH, biosensors.

  • Selection Criteria:

    • Accuracy, Range, Power consumption, Response time, Cost, Size, Calibration needs.

  • Common IoT Sensors:

    • DHT22: Temp/Humidity, ±0.5°C, digital.

    • MQ-2: Gas (LPG, smoke), analog.

    • PIR (HC-SR501): Motion, digital.

    • Ultrasonic (HC-SR04): Distance, 2–400 cm.

  • Quantization Error:

    • Error due to ADC converting continuous analog signal to discrete digital values.

    • Formula: $$\displaystyle Q_e = \pm \frac{FS}{2^{n+1}} $$

      • $FS$: Full-scale voltage range.

      • $n$: ADC resolution (bits).

    • Example: 10-bit ADC, 5V range → $$\displaystyle Q_e = \pm \frac{5}{2^{11}} = \pm 2.44\,\text{mV} $$.

Actuators

  • Types:

    • Mechanical: Motors (DC, stepper), relays, solenoids.

    • Soft: Silicone-based, flexible (wearables, robotics).

    • Shape Memory Polymer (SMP): Deforms with temperature/light, used in biomedical devices.

    • Pneumatic: Air pressure-driven, high force, used in industrial automation.

  • Four Selection Characteristics:

    1. Force/Torque Output.

    2. Response Time.

    3. Power Consumption.

    4. Precision/Accuracy.

  • Comparison:

    | Type | Advantages | Disadvantages | IoT Use Case | |----------------|---------------------------------|---------------------------------|-------------------------------| | Mechanical | High force, reliable | Noisy, bulky | Smart lock, valve control | | Soft | Flexible, safe for humans | Low force, complex control | Wearable haptics | | SMP | Remote stimulus response | Slow, temperature-sensitive | Self-deploying stents | | Pneumatic | High power-to-weight, explosion-proof | Needs air supply, leaks | Factory automation |

Microcontrollers & Interfacing

  • Basic MCUs:

    • Arduino (AVR): Easy prototyping, 5V, limited GPIO.

    • ESP32: Wi-Fi/Bluetooth, 3.3V, low-power modes.

    • AVR ATmega328: Standalone, low-cost.

  • Interfacing Fundamentals:

    | Interface | Type | Use Case | Pins (ESP32) | |---------------|----------------|---------------------------------------|-----------------------| | GPIO | Digital I/O | LED, button | Any digital pin | | SPI | Serial (sync) | High-speed (SD card, display) | SCK, MISO, MOSI, CS | | I2C | Serial (async) | Multiple sensors (same bus) | SCL, SDA | | UART | Serial (async) | GPS, GSM module | TX, RX | | ADC | Analog → Digital | Potentiometer, analog sensors | ADC1_0–ADC1_6 (18 ch)| | DAC | Digital → Analog | Audio output, analog control | DAC1, DAC2 (2 ch) |

Single-Board Computers (SBCs)

  • Raspberry Pi 3:

    • vs Desktop: ARM Cortex-A53 (1.2 GHz), 1 GB RAM, no internal storage (microSD), GPIO headers, lower power (~2.5W).

    • GPIO: 40-pin header, 3.3V logic, 17 GPIOs.

    • SPI/I2C: Enabled via config; multiple buses (SPI0, SPI1; I2C0, I2C1).

  • Other Platforms:

    • BeagleBone Black: PRU co-processors, more GPIO, industrial use.

    • Intel Galileo: x86 architecture, compatible with Arduino shields.

IoT Device Design Challenges

  • Power Management: Sleep modes, duty cycling, energy harvesting.

  • Form Factor: PCB size, antenna integration, enclosure.

  • Cost: BOM cost optimization.

  • Hardware-Software Co-design: Real-time OS selection, driver development, firmware updates OTA.

III. Communication Technologies & Protocols

Network & Data Link Layer Protocols

  • IEEE 802.15.4:

    • Features: Low-rate WPAN, 250 kbps (2.4 GHz), low power, star/mesh topologies, CSMA/CA.

    • Relation to IoT: Foundation for ZigBee, 6LoWPAN, Thread.

  • 6LoWPAN:

    • Role: Adaptation layer allowing IPv6 packets over IEEE 802.15.4.

    • How: Header compression, fragmentation.

    • vs IPv4/IPv6: Not a new IP version; enables IPv6 on constrained networks (address space for billions of devices).

  • ZigBee:

    • Architecture:

      
Physical/MAC (802.15.4) → Network/ Security (ZigBee) → Application (Profiles)
      • Components: Coordinator (forms network), Router (extends), End Device (sleeps).

    • Types: ZigBee PRO (mesh, large networks), ZigBee IP (IPv6), ZigBee RF4CE (remote control).

    • Features: Low power, mesh routing (AODV), 65,000 nodes, 128-bit AES.

  • Wireless Sensor Networks (WSN):

    • Concepts: Nodes (sensor+MCU+radio), Sink (data collector), multi-hop routing.

    • Topologies: Star (single-hop), Mesh (multi-hop), Tree (hierarchical).

    • Applications: Environmental monitoring, precision agriculture, health.

    • Relation to IoT: WSN is sensing infrastructure for IoT; IoT adds cloud/analytics/application layers.

[!TIP] Diagram Expected: ZigBee architecture (3-layer stack) and WSN topologies (star/mesh/tree).

Application Layer Protocols

  • MQTT:

    • Model: Publish-Subscribe (broker-centric).

    • Role: Lightweight, QoS levels (0,1,2), ideal for unreliable networks.

    • Example: Sensor publishes home/sensor/temp; mobile app subscribes.

    • QoS:

      • 0: At most once (fire-and-forget).

      • 1: At least once (acknowledged).

      • 2: Exactly once (4-step handshake).

  • CoAP:

    • Model: Request-Response (like HTTP), RESTful.

    • Use: Constrained networks (UDP, small headers), observe pattern for streaming.

    • Basic Operations: GET, POST, PUT, DELETE.

    • Same-network use: Devices communicate directly via multicast CoAP (e.g., coap://[ff02::fd] for all CoAP nodes).

  • AMQP:

    • Features: Binary, reliable, queuing, routing.

    • Components: Producer, Consumer, Queue, Exchange (direct/topic/fanout).

    • Message Attributes: content-type, correlation-id, priority.

    • Payload: Arbitrary binary.

    • Frame Types: 11 types (e.g., OPEN, BEGIN, ATTACH, FLOW, TRANSFER, DISPOSITION, CLOSE, DETACH, END, SASL-MECH, SASL-INIT).

  • XMPP:

    • Improves IoT: Built-in presence (device online/offline), federation (server-to-server), extensible (XEPs), real-time messaging.

  • SMQTT:

    • Secure MQTT: Uses symmetric encryption (AES) with key distribution via public key crypto (RSA). Messages encrypted before publishing.

  • WebSockets:

    • Role: Full-duplex, real-time communication over single TCP connection (replaces HTTP polling). Used for live dashboards, device control.

Other Communication Technologies

  • RFID:

    • Principles: Radio waves between tag & reader; passive (no battery), active (battery), semi-passive.

    • Features: Non-line-of-sight, unique ID, short range (cm–m).

    • IoT Integration: "Smart things" with RFID tags for tracking (supply chain, inventory).

  • NFC:

    • Technologies: 13.56 MHz, 4 cm range, peer-to-peer, card emulation.

    • IoT Applications: Device pairing (smartphone→IoT), contactless payments, access control.

  • Network Types (Physical Topologies):

    | Topology | Diagram | IoT Use | |--------------|--------------------------------------|---------------------------------| | Star |

    DiagramCANVAS: Central hub with radiating lines
    | Wi-Fi, BLE (gateway-centric) | | Mesh |
    DiagramCANVAS: Interconnected nodes
    | ZigBee, Thread (robust, scalable) | | Tree |
    DiagramCANVAS: Hierarchical branches
    | Industrial WSN (cluster-based) | | Bus |
    DiagramCANVAS: Single line with drops
    | Legacy wired systems |

[!TIP] Exam Focus: CoAP vs MQTT (UDP vs TCP, req-res vs pub-sub). SMQTT security mechanism (AES+RSA). RFID vs NFC range & use.

IV. Data Management, Cloud & Analytics

Data Analytics in IoT

  • Role: Transform raw sensor data → insights → actions (predictive maintenance, anomaly detection).

  • M2M vs IoT Analytics:

    • M2M: Batch processing, historical reports, closed systems.

    • IoT: Real-time streaming, predictive, cloud-based, open APIs.

  • Analytics Lifecycle:

    1. Collection: Sensors → gateway.

    2. Storage: Time-series DB (InfluxDB), cloud storage (S3).

    3. Processing: Stream (Apache Flink), batch (Spark).

    4. Visualization: Dashboards (Grafana), alerts.

Cloud Computing for IoT

  • Cloud Service Models:

    | Model | IoT Example | Control | |-----------|------------------------------------------|------------------------| | IaaS | AWS EC2 for custom IoT platform | High (OS, apps) | | PaaS | Azure IoT Hub (device mgmt., rules) | Medium (apps, data) | | SaaS | Salesforce IoT Cloud (pre-built analytics)| Low (configuration) |

  • Cloud Communication APIs:

    • REST/HTTP: Device-to-cloud (sensor data upload).

    • MQTT over Cloud: Pub-sub via cloud broker (AWS IoT Core).

    • CoAP: Constrained device-to-cloud.

  • IoT Platforms:

    • AWS IoT: Device Shadow, Greengrass (edge), Rules Engine.

    • Azure IoT: IoT Hub, Stream Analytics, Device Provisioning Service.

    • Google Cloud IoT: Core, Pub/Sub, Dataflow.

IoT Communication Models

  • Device-to-Device (D2D): Direct (BLE, ZigBee).

  • Device-to-Gateway (D2G): Sensor → gateway → cloud (common).

  • Device-to-Cloud (D2C): Direct IP (Wi-Fi/Ethernet).

  • Cloud-to-Cloud (C2C): Platform integration (AWS → Salesforce).

[!TIP] Diagram Expected: Four communication models with arrows between devices/gateway/cloud.

V. Security & Privacy in IoT

Need for Security in IoT

  • Why Critical:

    • Device Vulnerabilities: Weak passwords, unpatched firmware, physical access.

    • Data Sensitivity: Health data, home security, industrial control.

    • Attack Surfaces: Increased with each connected device (botnets like Mirai).

Security Models & Frameworks

  • Models:

    • Layered Security: Defense-in-depth (device, network, cloud, app).

    • Identity-Based: Device authentication (certificates, tokens).

    • Zero-Trust: "Never trust, always verify."

  • Requirements:

    • Confidentiality: Encryption (TLS, DTLS).

    • Integrity: HMAC, digital signatures.

    • Availability: DoS protection, redundancy.

    • Authentication: Mutual (device↔cloud).

    • Authorization: RBAC, scopes.

Vulnerabilities & Attacks

  • Kinds of Vulnerabilities:

    • Hardware: Debug interfaces (JTAG), side-channel attacks.

    • Software: Buffer overflows, hardcoded keys.

    • Network: Unencrypted traffic, open ports.

  • Attacks:

    • Application/Service Layer: Spoofing (fake sensor data), DoS/DDoS (Mirai), injection (CoAP flooding).

    • Physical: Tampering, theft.

    • Network: Sniffing, man-in-the-middle.

    • Protocol-Specific: Replay attacks (MQTT), fragmentation attacks (6LoWPAN).

Privacy Issues

  • Concerns:

    • Data Collection: Continuous monitoring (location, habits).

    • User Tracking: Profiling, behavioral advertising.

    • Data Sharing: Third-party sales without consent.

  • Mitigation:

    • Encryption: End-to-end.

    • Anonymization: Remove PII, use pseudonyms.

    • Access Control: Least privilege, audit logs.

    • Privacy by Design: Data minimization, user consent.

[!TIP] Exam Focus: List 3 attacks per layer (physical, network, app). Privacy vs security distinction: security protects data; privacy controls its use.

VI. Enablers, Case Studies & Advanced Topics

IoT Enablers Overview

  • Comprehensive List:

    1. Sensors/Actuators (data acquisition/action).

    2. Networks (Wi-Fi, LoRaWAN, 5G).

    3. Protocols (MQTT, CoAP, ZigBee).

    4. Hardware (MCUs, SBCs, gateways).

    5. Cloud Platforms (AWS, Azure).

    6. Analytics (ML, stream processing).

    7. Security (encryption, identity mgmt.).

  • Interdependencies Diagram:

    DiagramCANVAS: Central "IoT System" circle with 7 enablers as surrounding circles; arrows between all showing bidirectional dependencies (e.g., Sensors ↔ Protocols; Cloud ↔ Analytics; Security ↔ All).

Wireless Sensor Networks (WSN) Deep Dive

  • Concepts:

    • Node: Sensor + MCU + radio + battery.

    • Sink: Data collection point (gateway to internet).

    • Topology: Star, mesh, tree.

    • Routing Protocols: LEACH (cluster-based), AODV (on-demand), RPL (for 6LoWPAN).

  • Applications:

    • Environmental: Forest fire detection, air quality.

    • Agriculture: Soil moisture, crop monitoring.

    • Health: Patient vital signs.

  • Relation to IoT:

    • Integration: WSN provides sensing layer; IoT adds cloud/analytics/apps.

    • Difference: WSN focuses on data collection; IoT focuses on end-to-end services.

    • Convergence: WSN nodes become IoT devices with IP connectivity (6LoWPAN).

Case Study: Smart Home Automation

  • Design with Raspberry Pi:

    • Components:

      • Controller: Raspberry Pi 3 (runs Home Assistant).

      • Sensors: DHT22 (temp/humidity), PIR (motion), MQ-2 (gas).

      • Actuators: Relay module (lights/fan), servo (door lock).

      • Connectivity: Wi-Fi (Pi), ZigBee (sensors via CC2531 USB dongle).

      • Power: 5V/2.5A adapter for Pi; battery for sensors.

    • Neat Sketch:

      DiagramCANVAS: Raspberry Pi connected via USB to ZigBee coordinator; ZigBee sensors (DHT22, PIR) and actuators (relay) in rooms; Pi connected to Wi-Fi router → cloud → mobile app. Arrows show sensor data → Pi → cloud → app; app → cloud → Pi → actuator.

    • Use Cases:

      1. Lighting: Motion sensor → Pi → relay → light ON.

      2. Security: PIR triggers alert to mobile app.

      3. Climate Control: Temp > 30°C → Pi → fan relay ON.

Advanced & Emerging Topics

  • Software Defined Networking (SDN) in IoT:

    • Concept: Separate control plane (SDN controller) from data plane (switches).

    • Benefits: Centralized management, dynamic routing, security policies.

    • Maturity: Emerging in IoT (OpenFlow for WSN), not yet mainstream due to overhead.

  • Network Topologies & Connection Types (Diagrams as above).

  • Pneumatic Systems: Use compressed air; actuators (cylinders) for high-force, explosion-safe applications (factory automation).

  • IoT Applications:

    • Industrial: Predictive maintenance (vibration sensors).

    • Smart Cities: Traffic monitoring, waste management.

    • Agriculture: Precision farming (soil sensors, automated irrigation).

  • Future Trends:

    • AI at Edge: TinyML on microcontrollers.

    • 5G for IoT: URLLC (ultra-reliable low-latency), mMTC (massive machine-type comms).

    • Blockchain for Security: Decentralized device identity, audit trail.

[!TIP] Exam Focus: Smart home sketch must include Pi, sensors, actuators, connectivity. SDN: explain separation of control/data planes. 5G vs 4G for IoT: latency, capacity.

Final Exam Strategy:

  1. Definitions First: Start answers with clear definitions (e.g., "IoT is...").

  2. Diagrams: Label all components (ZigBee layers, WSN topologies, smart home).

  3. Compare/Contrast: Use tables for M2M vs IoT, MQTT vs CoAP, actuator types.

  4. Formulas: Box key formulas (quantization error, load factor if in context).

  5. Examples: Relate every concept to real IoT use cases (e.g., "6LoWPAN enables smart agriculture sensors").

Boxed Key Formulas:

  • Quantization Error: $$\displaystyle \boxed{Q_e = \pm \frac{FS}{2^{n+1}}} $$

  • IoT Definition: $$\displaystyle \boxed{\text{IoT} = \text{Things} + \text{Internet} + \text{Analytics}} $$

Go to where you left off?

Quick Add to Notes

Save questions, your own notes and screenshots into notes filed by unit. It takes a free account.

Create free account

Have an account? Log in