Internet of Things (CE-703 (A)) - Unit 5 Short Notes

UNIT 5: INTERNET OF THINGS (CE-703 A) - EXAM-FOCUSED NOTES

1. IoT FUNDAMENTALS & ARCHITECTURE

IoT Definition & Core Concepts

• IoT Definition: A system of interrelated computing devices, mechanical and 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.

• IoT vs. M2M:

  • M2M (Machine-to-Machine): Point-to-point communication between two machines. Standalone, closed-loop, often proprietary. Focus is on data collection.

  • IoT: Network of 'things' connected to the internet/cloud. Open, scalable, standardized protocols. Focus is on data, analytics, and services.

  • Evolution Rationale: M2M is a subset/enabler of IoT. IoT adds cloud integration, big data analytics, and a broader ecosystem of applications and stakeholders.

IoT Architectural Frameworks

• IoT Reference Architectures (e.g., IETF, Cisco, IoT-A): Provide standardized blueprints defining components, interfaces, and protocols. Ensure interoperability and guide implementation.

• IoT Levels/System Types:

  | Level | Description | Example | | :--- | :--- | :--- | | 1 | Single device, no internet connectivity. | Standalone sensor with local display. | | 2 | Single device connected to internet. | Smart thermostat connected via Wi-Fi. | | 3 | Local network of devices (gateway-based). | Smart home hub connecting lights, sensors. | | 4 | Global system with cloud integration. | Smart city platform with city-wide sensors & cloud analytics. |

• IoT Planes & Enablers: Interdependent layers like Management Plane (device mgmt.), Communication Plane (networks), Application Plane (services), Security Plane. Enablers (e.g., cloud, big data, AI) sit across these planes.

• Service-Oriented Architecture (SOA) for IoT: Principles: Loose coupling, service reusability, standardized contracts. Challenges: Resource constraints on devices, dynamic service discovery, real-time requirements.

• Logical vs. Physical Design:

  • Logical Design: Functional view. Defines services, processes, data models, APIs. (e.g., "data acquisition service," "alert notification process").

  • Physical Design: Implementation view. Defines hardware (sensors, actuators, gateways), software stacks, network protocols, physical connectivity.

IoT Characteristics & Functional Blocks

• Core Characteristics: Connectivity, Intelligence (data processing), Scalability, Adaptability, Architecture heterogeneity.

• Building Blocks: Things (sensors/actuators), Gateways (local intelligence, protocol translation), Network (communication), Cloud/Data Center (storage, analytics), Applications (user interface, business logic).

[!TIP] Exam Focus: Differentiating IoT Levels (1-4) and Logical vs. Physical design are very high frequency. Be ready to draw/explain with examples.

2. DEVICE LAYER: HARDWARE & SENSING

Sensors & Actuators

• Sensor Features & Characteristics:

  • Accuracy: Closeness to true value.

  • Precision: Repeatability of measurements.

  • Range: Min & Max measurable values.

  • Resolution: Smallest detectable change.

  • Error: Difference between measured & true value.

• Common Commercially Available Sensors: Temperature (DHT11, DS18B20), Humidity, PIR (motion), Gas (MQ series), Proximity (Ultrasonic, IR), Pressure, Light (LDR).

• Actuators:

  • Types:

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

    2. Soft: Pneumatic muscles, elastomers.

    3. Shape Memory Polymer (SMP): Change shape with temperature/light.

  • Selection Characteristics: Force/torque, speed, precision, power source, environment, cost.

• Pneumatic Actuation: Uses compressed air to produce motion. Components: compressor, valves, cylinders. Advantages: Clean, safe (spark-free), high power-to-weight. Disadvantages: Requires air supply, less precise.

Microcontrollers & Single-Board Computers

• Basic Microcontrollers (Arduino, ESP32): Integrated CPU, memory, I/O pins. Low cost, low power. Programmed via IDE. Interfacing: Connect sensors/actuators via digital/analog pins, communication protocols (I2C, SPI, UART).

• Raspberry Pi: Full-fledged SBC with OS (Linux). Architecture: SoC (Broadcom), GPIO pins, USB, HDMI, Ethernet. GPIO: General Purpose Input/Output pins for digital/analog (via ADC) signals.

  • SPI (Serial Peripheral Interface): Synchronous, full-duplex, master-slave. 4 wires (MOSI, MISO, SCLK, CS). Fast, for short distances.

  • I2C (Inter-Integrated Circuit): Synchronous, multi-master, multi-slave. 2 wires (SDA, SCL). Address-based, for moderate speeds.

• Comparison with Desktop: Pi has lower processing power, no BIOS, boots from SD card, designed for embedded control vs. general-purpose computing.

• Quantization Error: Error in Analog-to-Digital Conversion (ADC). Difference between actual analog input and its digital representation. Caused by finite resolution (bits). Formula: Max error = ±½ LSB (Least Significant Bit).

Identification Technologies

• RFID (Radio Frequency Identification):

  • Principle: Uses electromagnetic fields to automatically identify and track tags attached to objects.

  • Components: Tag (passive/active, chip+antenna), Reader (transmitter/receiver), Antenna, Middleware/Backend System.

  • Features: Non-line-of-sight, fast read, unique ID, durable.

  • Link to IoT: Provides unique digital identity ('thing') for physical objects, enabling tracking, inventory, and automation.

• NFC (Near Field Communication): Short-range (~10 cm) wireless tech. Based on RFID. Applications: Contactless payments, data exchange, device pairing.

[!TIP] Exam Focus: Sensor characteristics and RFID features are very high. Raspberry Pi interfaces (SPI/I2C) and Quantization error are medium. Be precise with definitions and formulas.

3. COMMUNICATION & NETWORKING PROTOCOLS

Low-Power Wireless Network Standards

• IEEE 802.15.4: Foundation for many IoT protocols (ZigBee, 6LoWPAN). Defines PHY (O-QPSK, DSSS) and MAC (CSMA/CA, beacon-enabled) layers for low-rate, low-power WPANs. Frame structure: Preamble, PHY Header, MAC Header, Payload, CRC.

• 6LoWPAN (IPv6 over Low-Power WPAN):

  • Functionality: Adaptation layer enabling IPv6 packets to travel over IEEE 802.15.4 networks. Handles fragmentation/ reassembly of IPv6 packets (max 127 bytes) to fit 802.15.4 frame.

  • Differentiation from IPv4/IPv6: Not a replacement. It's an adaptation layer for constrained links. IPv6 provides vast address space (~3.4×10³⁸ addresses), crucial for IoT scalability, unlike IPv4's limited space.

• ZigBee:

  • Protocol Stack: Based on IEEE 802.15.4 (PHY/MAC). Adds Network (NWK) and Application (APL) layers.

  • Architecture:

    • Coordinator: Forms network, stores network info, may be gateway.

    • Router: Extends network range, routes data, can join children.

    • End Device: Leaf node, talks only to parent (coordinator/router), sleeps to save power.

  • Types: ZigBee PRO (standard, for large networks), ZigBee IP (for IPv6), ZigBee RF4CE (for consumer electronics).

Constrained Application Layer Protocols

• CoAP (Constrained Application Protocol):

  • Model: RESTful request-response (like HTTP). Uses Confirmable (CON) and Non-Confirmable (NON) messages for reliability/overhead trade-off.

  • Use in Constrained Networks: Lightweight (header ~4 bytes), UDP-based (no connection overhead), supports multicast. Ideal for M2M between devices on same local constrained network (e.g., sensor to gateway).

  • Basic Operations: GET (retrieve), POST (create), PUT (update), DELETE (remove). Resources identified by URIs (e.g., /sensors/temp).

• MQTT (Message Queuing Telemetry Transport):

  • Model: Publish/Subscribe. Clients publish to Topics (e.g., home/livingroom/temp). Broker filters and delivers to subscribers.

  • Role in IoT: Lightweight, efficient for unreliable networks, supports many clients. Decouples producers/consumers.

  • WebSockets: MQTT can run over WebSockets for browser-based clients, enabling real-time web dashboards.

• AMQP (Advanced Message Queuing Protocol):

  • Features: Binary, wire-level protocol. Components: Exchanges (receive producer msgs), Queues (store msgs), Bindings (rules linking exchange to queue).

  • Message Attributes & Payload: Message has properties (content-type, priority, timestamp) and body (payload). Enables rich messaging.

  • Frame Types: Protocol frames for different functions: OPEN, BEGIN, ATTACH, SEND, FLOW, DISPOSITION, CLOSE, END.

• XMPP (Extensible Messaging and Presence Protocol): XML-based, pub/sub. Improves IoT Services: Built-in presence (device status), federation (inter-domain comm.), security (TLS/SASL), extensibility via XEPs (e.g., sensor data).

• SMQTT (Secure MQTT): MQTT with security layer. Uses cryptography (e.g., AES) to encrypt payload and signature for integrity/authentication. Prevents eavesdropping/tampering.

Network Types & Topologies

• Classification:

  • Physical Topologies: Star (all to hub), Mesh (peer-to-peer, multi-hop), Tree (hierarchical).

  • Connection Types: WPAN (Bluetooth, ZigBee - short range), WLAN (Wi-Fi - medium range), LPWAN (LoRaWAN, NB-IoT - long range, low power).

• Wireless Sensor Networks (WSN):

  • Architecture: Many sensor nodes (sense, process, transmit) → Sink node → Base station → User/Cloud.

  • Characteristics: Self-organizing, multi-hop, resource-constrained, data-centric.

  • Relation to IoT: WSN is the sensing/data acquisition layer of IoT. Example: Soil moisture sensors in a field (WSN) sending data to a cloud farm management platform (IoT).

• IoT LAN Development Issues: Interference (Wi-Fi, Bluetooth), power management for battery nodes, scalability of mesh networks, security in open protocols, gateway selection and placement.

[!TIP] Exam Focus: 6LoWPAN vs IPv6, ZigBee architecture, CoAP request-response, MQTT pub/sub, AMQP components/frames are very high. Draw ZigBee topologies and AMQP exchange/queue diagrams.

4. MIDDLEWARE, GATEWAYS & CLOUD INTEGRATION

IoT Gateways

• Functionality & Role:

  1. Protocol Translation: Convert between device protocols (ZigBee, BLE) and internet protocols (HTTP, MQTT).

  2. Data Filtering & Aggregation: Pre-process raw sensor data, reduce cloud traffic.

  3. Edge Processing/Intelligence: Local analytics, rule execution (e.g., "if temp>30, turn on fan").

  4. Security: Firewall, encryption, device authentication.

  5. Device Management: Onboarding, monitoring, OTA updates.

• IoT Ecosystem End-to-End: Things (sensors) → Gateway (local network, protocol conversion) → Internet/Cloud (storage, analytics) → Application (user dashboard, alerts).

Cloud Computing for IoT

• Usefulness of Cloud for IoT:

  • Scalability: Handle millions of devices/data points.

  • Storage: Massive, durable storage for historical data.

  • Analytics: Powerful compute for big data, ML/AI.

  • Cost-Effective: Pay-as-you-go, no upfront infrastructure.

• Cloud Communication APIs: RESTful APIs (HTTP methods: GET/POST/PUT/DELETE) are standard. Devices/gateways send data to cloud endpoints (e.g., POST /api/devices/{id}/telemetry). Cloud returns JSON/XML.

• Cloud Service Models in IoT Context:

  • IaaS (Infrastructure): Rent VMs, storage (e.g., AWS EC2). User manages OS, runtime, apps.

  • PaaS (Platform): Platform for IoT app dev (e.g., AWS IoT Core, Azure IoT Hub). Manages infrastructure, provides device mgmt., data ingestion.

  • SaaS (Application): Ready-to-use IoT apps (e.g., Salesforce IoT Cloud, predictive maintenance dashboards).

Data Analytics in IoT

• Role & Importance:

  • Descriptive: What happened? (Dashboards, reports).

  • Diagnostic: Why did it happen? (Root cause analysis).

  • Predictive: What will happen? (Forecasting, anomaly detection).

  • Prescriptive: What should we do? (Optimization, recommendations).

• Challenges: Volume, Velocity, Variety (3Vs) of data. Real-time processing needs. Data quality (noise, missing values). Privacy concerns. Integration with legacy systems.

[!TIP] Exam Focus: Gateway functionality and Cloud usefulness are very high. Be ready to explain the ecosystem flow and differentiate IaaS/PaaS/SaaS with IoT examples.

5. SECURITY & PRIVACY IN IoT

Need for Security in IoT

• Unique vulnerabilities: Resource constraints limit strong crypto, physical accessibility of devices, heterogeneous networks, large attack surface, often lack of updates.

IoT Security Models & Frameworks

• Models focus on layered security: Device/Hardware, Network, Cloud/Application.

• Frameworks (e.g., NIST, IoT Security Foundation) provide guidelines: device identity, secure boot, encrypted comms, access control, privacy by design.

Vulnerabilities & Attacks

• Observed Vulnerabilities: Hardcoded passwords, unencrypted comms, insecure web interfaces, lack of firmware updates, poor physical security.

• Attacks on Application/Service Layer:

  • Spoofing: Faking device/user identity.

  • Tampering: Modifying data in transit/at rest.

  • Information Disclosure: Leaking sensitive data.

  • Denial-of-Service (DoS): Overwhelming service/device.

  • Injection: SQL, command injection via app inputs.

• General Attacks on IoT Systems:

  • Physical: Tampering, side-channel attacks.

  • Network: Sniffing, MITM, routing attacks.

  • Application: As above.

  • Cryptographic: Key extraction, weak crypto.

Security & Privacy Issues & Mitigation

• Issues: Data privacy (user behavior tracking), device hijacking (botnets like Mirai), safety risks (medical/industrial).

• Mitigation Strategies:

  • Device: Secure boot, HW security modules, regular patching.

  • Network: Encryption (TLS/DTLS), firewalls, segmentation.

  • Cloud/App: Strong auth (OAuth2, certs), input validation, audit logs.

  • Privacy: Data minimization, anonymization, user consent.

[!TIP] Exam Focus: Why security needed, various attacks (especially app layer), and mitigation are very high. Use examples like Mirai botnet for DoS.

6. ENABLING TECHNOLOGIES & ADVANCED TOPICS

• IoT Enablers: Key technologies that make IoT feasible: Cloud Computing, Big Data Analytics, AI/ML, 5G, Edge Computing, Blockchain (for trust).

• Software Defined Networking (SDN): Separates control plane (centralized controller) from data plane (switches). Concept for IoT: Enables flexible, programmable network management for dynamic IoT topologies. Maturity: Emerging for IoT; challenges in scalability for massive device networks and controller placement.

• WebSockets: Provides full-duplex communication over single TCP connection. Role in IoT: Enables real-time, bidirectional data push from server to client (e.g., live dashboard updates), avoiding HTTP polling overhead.

7. APPLICATIONS & CASE STUDIES

Smart Home Automation (Design with Raspberry Pi)

• System Design:

  • Central Hub: Raspberry Pi (runs OS, broker like Mosquitto, web server).

  • Sensors: DHT11 (temp/humidity), PIR (motion), LDR (light).

  • Actuators: Relays (for lights/fans), servo motors (for door locks).

  • Connectivity: Sensors → Pi via GPIO/I2C/SPI. Pi → Cloud via Wi-Fi/Ethernet.

  • User Interface: Mobile app/Web page (subscribes to MQTT topics from Pi).

• Neat Sketch Requirement:

  
  [Sensors] --> (GPIO/I2C) --> [Raspberry Pi Hub] --(Wi-Fi)--> [Cloud/Internet] <--(App)--> [User]
  
                    |                                     |
  
                [Relays/Actuators]                 [MQTT Broker on Pi]
  
  

Other Application Domains

• Smart Agriculture: Soil moisture sensors → irrigation control (actuators) → cloud analytics for water optimization.

• Smart City: Smart bins (fill-level sensors → optimized collection routes), traffic sensors → adaptive signals.

• Industrial IoT (IIoT): Machine vibration sensors → predictive maintenance.

Case Study Analysis & Cloud Integration Challenges

• Example Case: Smart Agriculture.

  • Challenge: Integrating field sensors (LoRaWAN) with cloud (AWS IoT). LoRaWAN gateway sends data to cloud via MQTT, but field devices have intermittent connectivity.

  • Solution Approach: Use edge gateway (Raspberry Pi) with local storage to buffer data during network outage, then sync when connection resumes. Implement store-and-forward logic.

[!TIP] Exam Focus: Smart Home design with sketch is very high. Practice drawing the block diagram. For case studies, focus on challenges (connectivity, power, scale) and solutions (edge computing, protocol choice).

8. DESIGN & IMPLEMENTATION CHALLENGES

• Challenges & Requirements of IoT Devices:

  | Challenge | Requirement | | :--- | :--- | | Power | Battery-operated, low-power modes, energy harvesting. | | Processing | Sufficient for local tasks, but constrained (CPU, memory). | | Cost | Ultra-low cost for mass deployment. | | Size | Miniaturization. | | Security | Secure boot, encryption, key management within constraints. | | Connectivity | Reliable, low-power wireless (LPWAN, BLE). | | Durability | Operate in harsh environments (temp, humidity). |

• Issues in IoT LAN Development: Interference, network scalability, device interoperability (different protocols), gateway bottleneck, physical deployment (placement of nodes/gateways).

• Security Model Selection & Implementation: Choose model based on risk assessment. For critical systems (medical), use strict models (zero-trust). For consumer IoT, balance security with usability/cost. Implementation: Start with device identity (certificates), enforce encryption in transit (TLS), use gateway as security perimeter.

[!TIP] Exam Focus: Device challenges table and IoT LAN issues are very high. Link challenges to specific application domains (e.g., power for remote agriculture sensors).