UNIT 1

1. Concept of IoT and Its Significance

Definition:

• IoT = network of physical objects embedded with sensors, software, actuators, and
connectivity

• Enables devices to collect and exchange data over the internet

• No human-to-human or human-to-computer interaction required

• Term coined by Kevin Ashton in 1999

• Connects the physical world to the digital world

Significance:

• Enables automation, remote monitoring, and intelligent decision-making at scale

• Reduces human error and improves efficiency

• Enables predictive maintenance in industries

• Generates large volumes of data that fuel AI and analytics

• Reshapes industries: healthcare, agriculture, manufacturing

• Smart cities use IoT for traffic, waste, and energy management

• Enables Industry 4.0 — cyber-physical systems in manufacturing

• Economic impact projected in trillions of dollars globally

2. Communication Models in IoT

Device-to-Device (D2D):

• Two or more devices communicate directly without an intermediary server

• Uses protocols like Zigbee or Bluetooth

• Low-latency and efficient for local control

• Limited in range and scalability

• Example: smart locks and lighting systems in smart homes

Device-to-Cloud:

• Devices connect directly to a cloud service provider

• Most common IoT communication model

• Offers scalability and remote access

• Depends heavily on internet connectivity

 • Example: smart thermostat sending data to AWS IoT

Device-to-Gateway:

• Devices communicate with an intermediate gateway

• Gateway aggregates data and forwards to cloud/server

• Necessary when devices use low-power protocols (Zigbee, Z-Wave)

• Gateway acts as a protocol translator

• Example: industrial IoT sensors connecting via gateway

Back-End Data Sharing:

• Multiple cloud platforms share and analyze data from IoT devices

• Supports interoperability across organizations

• Used in smart city infrastructure

• Example: different city agencies sharing real-time sensor data

Comparison Table:

Model Direct Internet? Scalability Latency Use Case

D2D No Low Very Low Home automation

D2Cloud Yes High Medium Wearables, tracking

D2Gateway No (via GW) Medium Low-Medium Industrial IoT

Back-End Sharing Yes Very High High Smart cities

3. Logical Design of IoT — Functional Blocks

Device Block:

• Physical "thing" with sensors and actuators

• Collects real-world data (temperature, motion, pressure)

• Performs physical actions (open valve, turn on motor)

Communication Block:

• Handles data transmission between devices and system

• Includes hardware interfaces: Wi-Fi, Zigbee, LTE

• Includes protocols: MQTT, CoAP, HTTP

Services Block:

• Middleware services for device management

 • Handles data publishing/subscribing

• Enables service discovery

• Abstracts hardware complexity from applications

Management Block:

• Covers operational aspects of the system

• Includes fault management, configuration, performance monitoring

• Handles security operations

• Uses protocols like NETCONF and YANG

Security Block:

• Ensures authentication and authorization

• Provides data integrity and encryption

• Cross-cutting concern embedded throughout all layers

Application Block:

• Top-level user-facing layer

• Data is visualized, analyzed, and acted upon

• Includes dashboards, mobile apps, analytics platforms

4. Major Challenges Facing IoT Implementation

Security and Privacy:

• IoT devices are resource-constrained and cannot run heavy encryption

• Become easy entry points for cyberattacks

• Mirai botnet attack (2016) hijacked millions of IoT devices

• Smart devices collect deeply personal data raising privacy concerns

Interoperability:

• Thousands of vendors use different protocols and platforms

• No universal standard exists across the ecosystem

• Devices from different vendors may not communicate with each other

• Fragments the overall IoT ecosystem

Scalability:

• Billions of devices coming online simultaneously

• Traditional centralized architectures struggle to keep up

 • Networks, cloud platforms, and data pipelines must scale massively

Power and Energy Constraints:

• Many IoT devices are battery-operated

• Must run for months or years without recharging

• Energy-efficient protocols and hardware design are critical

Data Management:

• IoT generates massive volumes of data continuously

• Storing, processing, and extracting insights in real time is challenging

• Requires robust big data infrastructure

Reliability and Connectivity:

• Network connectivity may be intermittent in remote/industrial settings

• Systems must be resilient and capable of offline operation

Regulatory and Legal Issues:

• Data sovereignty and compliance with GDPR adds complexity

• Liability for device failures is legally unclear

• Global deployments face different laws in different countries

5. Applications of IoT Across Different Sectors

Smart Home:

• Automated lighting and smart thermostats (e.g., Nest)

• Security cameras and voice assistants (e.g., Alexa)

• Connected appliances improve comfort and energy efficiency

Healthcare:

• Wearable devices monitor heart rate, glucose, oxygen levels

• Remote patient monitoring reduces hospital admissions

• Smart pills can track drug ingestion in real time

Agriculture (Precision Farming):

• Soil moisture sensors optimize irrigation

• Drone-based crop monitoring improves yield

• Automated systems reduce water and fertilizer waste

Manufacturing (IIoT):
 • Predictive maintenance of machinery reduces downtime

• Real-time quality control on factory floors

• Robot coordination and automation enabled

Smart Cities:

• IoT sensors manage traffic lights intelligently

• Monitor air quality and environmental conditions

• Optimize waste collection routes and public lighting

Transportation and Logistics:

• Fleet tracking and cold chain monitoring

• Predictive vehicle maintenance

• V2V (vehicle-to-vehicle) communication for safety

Energy Management:

• Smart meters track consumption in real time

• Smart grids balance supply and demand dynamically

• Better integration of renewable energy sources

Retail:

• Smart shelves track inventory automatically

• Beacons personalize in-store customer experiences

• Connected checkout systems reduce queues

6. Physical Architecture of an IoT System

Perception Layer (Device Layer) — Bottom Layer:

• Contains physical sensors: temperature, humidity, motion, GPS

• Contains actuators: motors, relays, valves

• Uses embedded processors: Arduino, Raspberry Pi, ARM Cortex

• Directly interacts with the physical environment

Network Layer (Communication Layer) — Middle Layer:

• Transmits data from devices to processing systems

• PAN technologies: Bluetooth, Zigbee (short range)

• LAN technologies: Wi-Fi, Ethernet (medium range)

• WAN technologies: 4G/5G, LoRa, NB-IoT (long range)

 • Gateways bridge short-range protocols to internet protocols

Application Layer (Processing/Service Layer) — Top Layer:

• Stores, processes, and presents data to users

• Includes cloud platforms: AWS IoT, Azure IoT Hub, Google Cloud IoT

• Includes databases, analytics engines, and dashboards

• User-facing applications: mobile apps, APIs, web portals

Fog/Edge Computing Layer (Modern Addition):

• Sits between network and application layers

• Processes data closer to the source

• Reduces latency and bandwidth requirements to cloud

7. Key Characteristics of IoT Devices

• Connectivity: Must connect to a network directly or via gateway using Wi-Fi, Bluetooth,
Zigbee, cellular, etc.

• Sensing/Actuation: Can sense environmental data through sensors or perform actions via
actuators, or both

• Heterogeneity: Extremely diverse in hardware, software, OS, and communication protocols

• Resource Constraints: Limited CPU power, memory, storage, and battery capacity

• Dynamic Nature: Devices join/leave networks constantly; operational context shifts

continuously

• Intelligence: Many modern devices embed on-device ML (TinyML) for local decision-making

• Scalability: Infrastructure must handle millions or billions of devices simultaneously

• Safety: Must meet safety standards, especially in industrial, medical, and automotive
contexts

UNIT 2

1. IoT and M2M Communication

M2M (Machine-to-Machine):

• Direct communication between devices without human intervention

• Uses any communication channel — wired or wireless

• Predates IoT; used in industrial SCADA and telemetry systems

 • Typically point-to-point using proprietary protocols

• Operates over closed, private networks

IoT:

• Builds upon and extends M2M concepts

• Open, IP-based, internet-connected ecosystem

• Devices interact with each other, cloud services, and humans

• Adds intelligence, interoperability, and global connectivity

Key Relationship:

• M2M is a subset and predecessor of IoT

• IoT adds cloud platforms, analytics, and human-machine interfaces on top of M2M principles

2. Architecture of SDN

Overview:

• SDN = Software-Defined Networking

• Decouples the control plane (decision-making) from the data plane (packet forwarding)

• Centralizes network intelligence in software

Application Layer (Top):

• SDN applications define network behavior

• Includes: load balancing, firewalls, traffic engineering

• Communicates with controller via Northbound Interface (NBI)

• NBI is typically REST-based API

Control Layer (Middle):

• Contains the SDN Controller — the "brain" of the network

• Maintains a global view of the network topology

• Programs forwarding behavior of all switches

• Examples: OpenDaylight, ONOS, Floodlight

• Communicates downward via Southbound Interface (SBI)

Infrastructure Layer (Bottom):

• Physical or virtual network devices (switches, routers)

• Forwards traffic based on rules from the controller

• OpenFlow is the most common SBI protocol

 • Controller populates flow tables in switches via OpenFlow

East-Westbound Interfaces:

• Allow communication between multiple SDN controllers

• Enables federated control across different network domains

3. Key Features and Applications of M2M

Key Features:

• Autonomous operation — no human in the loop required

• Real-time data collection from remote devices

• Uses embedded SIM cards or dedicated radio modules

• Low-power consumption for remote/battery-operated deployment

• Uses both licensed spectrum (2G/3G/4G) and unlicensed (Wi-Fi, Zigbee)

• Operates in vertical silos — each application has dedicated infrastructure

• Uses proprietary protocols unlike IoT's open standards

Applications:

• Utilities: Smart metering for electricity, water, gas — eliminates manual reading

• Transportation: Vehicle tracking and fleet management systems

• Healthcare: Remote patient monitoring via connected medical devices

• Vending Machines: Operators remotely check stock levels and receive fault alerts

• Security Systems: Alarm systems automatically connected to monitoring centers

• Industrial: SCADA systems for monitoring and controlling industrial processes

4. Rule Replacement in SDN

Background:

• Network devices (switches) maintain flow tables with rules for handling packets

• Each rule has: match field (source IP, destination port), action (forward/drop/modify),
priority, and timeout

• Hardware switches have limited TCAM memory for flow tables

Why Rule Replacement is Needed:

• When flow table reaches capacity, old rules must be removed to make space for new ones

• Similar to cache replacement in computer architecture

 • Critical in IoT-SDN because IoT generates enormous numbers of short-lived flows

Replacement Strategies:

• LRU (Least Recently Used): Remove the rule with the oldest last-access timestamp

• LFU (Least Frequently Used): Remove the rule accessed least often

• TTL-based Expiry: Rules expire automatically after a timeout period defined at installation

Impact of Poor Rule Replacement:

• Increases flow setup latency — controller must be contacted for every cache miss

• Degrades overall network performance

• Especially harmful in high-traffic IoT environments

5. Differences Between IoT and M2M

Dimension M2M IoT

Communication Point-to-point Many-to-many, cloud-centric

Network Proprietary, closed IP-based, open internet

Architecture Vertical silos Horizontal platform

Intelligence Device-level only Cloud + device + edge

Scalability Limited Virtually unlimited

Standards Proprietary Open (MQTT, CoAP, REST)

Human Interaction Minimal Significant (apps, dashboards)

Data Usage Operational data only Big data, analytics, AI

Examples SCADA, telemetry Smart home, smart city

Summary:

• M2M focuses on connecting machines for specific tasks using closed systems

• IoT creates an open, intelligent, scalable ecosystem far beyond M2M's scope

• M2M is considered a subset and predecessor of IoT

6. Requirements of Network Operators in Managing IoT

Device Onboarding and Provisioning:

• Millions of devices need zero-touch provisioning

• Devices must connect, authenticate, and receive configuration automatically

 • No manual intervention should be required at scale

Remote Configuration and Updates:

• Devices often deployed in inaccessible locations

• Operators must push firmware updates and security patches over-the-air (OTA)

• Configuration changes must be deployable remotely

Monitoring and Fault Management:

• Continuous monitoring of device health, connectivity, and performance

• Automated fault detection and alerting systems required

• Reduces operational costs significantly

Security Management:

• Enforce strong authentication mechanisms

• Manage certificates and cryptographic keys

• Detect anomalous behavior indicating device compromise

• Quickly isolate and quarantine infected devices

Quality of Service (QoS):

• Different IoT applications have different latency/bandwidth needs

• Real-time control systems need low latency

• Bulk sensor data can tolerate delays

• Operators must prioritize traffic intelligently

Scalability:

• Management systems must handle billions of endpoints

• Requires distributed, cloud-native management architectures

Interoperability:

• Operators manage heterogeneous device ecosystems

• Standards-based protocols required: NETCONF, YANG, SNMP, TR-069

7. Roles of SDN and NFV in IoT

SDN in IoT:

• Provides programmable, centralized control to IoT networks

• Controller dynamically adjusts routing, security policies, and bandwidth allocation

• Network adapts rapidly to changing device states

 • Enables network slicing — logically separate virtual networks on shared physical
infrastructure

• Isolates different IoT applications for security and performance

• Single control point manages entire IoT network instead of device-by-device configuration

NFV in IoT:

• NFV = Network Function Virtualization

• Replaces dedicated hardware appliances with software on commodity servers

• Virtual functions include: firewalls, NAT gateways, load balancers, protocol converters

• IoT gateways and edge nodes use NFV for protocol translation and data aggregation

• Functions can be deployed, scaled, and updated in software without hardware changes

• Example: virtual firewall instantiated on demand when a new IoT device type joins the
network

Combined Impact of SDN + NFV:

• Together they enable "programmable infrastructure" for large-scale IoT

• Dynamic, automated, scalable, and cost-efficient network management

• Overcomes limitations of traditional hardware-centric networking

• Essential for managing billions of heterogeneous IoT devices efficiently

8. IoT Systems Management: NETCONF and YANG (10M)

Importance of IoT Systems Management:

• Encompasses monitoring, configuring, maintaining, and securing IoT devices

• At millions of devices, manual management is impossible

• Automated, standardized management is absolutely necessary

• Poor management leads to security vulnerabilities (unpatched devices)

• Poor management causes service degradation and operational chaos

• Must address: configuration, fault, performance, security, and lifecycle management

NETCONF (Network Configuration Protocol):

• IETF-standardized protocol — RFC 6241

• Designed for installing, manipulating, and deleting network device configurations

• Operates over SSH for secure transport

• Uses XML for data encoding

 • Transactional in nature — ensures configuration consistency

NETCONF Operations:

• get — retrieve operational state data

• get-config — retrieve configuration data

• edit-config — modify configuration

• copy-config — copy one datastore to another

• delete-config — delete a configuration datastore

• lock/unlock — prevent concurrent modification

• commit — apply candidate configuration to running state

NETCONF Datastores:

• Running — currently active configuration on the device

• Candidate — proposed configuration being edited/tested

• Startup — configuration loaded when device boots

Benefits of NETCONF in IoT:

• Remotely configure gateways, edge routers, and IoT devices

• Transactional commits prevent partial or broken configurations

• Standardized interface works across multiple vendors

YANG (Yet Another Next Generation):

• Data modeling language — RFC 6020 and RFC 7950

• Models the configuration and operational state data that NETCONF manages

• Defines structure, syntax, and semantics of network data

• Human-readable and machine-processable format

• Data is modeled as a tree of nodes

Key YANG Constructs:

• module — top-level container for a YANG model

• container — grouping of related nodes

• list — sequence of entries identified by a key

• leaf — single data value (like a variable)

• leaf-list — sequence of single values

• typedef — reusable type definitions

 • RPC — remote procedure call definitions

• notification — event-driven communication definitions

Example YANG Model for IoT Sensor:

• Container: "sensor"

• Leaves: sensor-id, location, current-temperature, threshold, alert-enabled

• Operator uses NETCONF to read temperature, update threshold, or enable alerts

Significance of NETCONF + YANG in IoT:

• Provides vendor-neutral, standardized interface for IoT device management

• Vital for interoperability in heterogeneous IoT ecosystems

• YANG models can be shared publicly between vendors

• Management systems can configure devices from any vendor using published YANG models

• Major step toward plug-and-play IoT vision

• OpenConfig and IETF YANG modules accelerating standardization

• YANG models consumed by SDN controllers for automated management

• Enables intent-based management — operators specify desired outcomes, system figures

out configuration

• Key enabler for autonomous IoT network operations