Introduction to IoT

Introduction to IoT

• The Internet of Things (IoT) is a network connecting physical objects embedded with sensors, software, and network connectivity for data exchange.
• An example of IoT is a smart thermostat adjusting temperature based on real-time weather data.
• Kevin Ashton coined the term "Internet of Things" in 1999.
• The first IoT device was a modified Coca-Cola vending machine at Carnegie Mellon University in the early 1980s, reporting stock levels remotely.
• Key characteristics of IoT include connectivity via Wi-Fi, Bluetooth, or cellular networks.
• IoT has automation capabilities without human intervention.
• Sensors enable data collection and analysis for decision-making.

Applications and Examples

• In smart homes, IoT enables devices like Amazon Alexa, Google Nest, and smart lighting.
• Wearable devices like Fitbit and Apple Watch utilize IoT.
• In healthcare, remote patient monitoring uses IoT-enabled medical devices.
• Industrial IoT (IIoT) involves smart manufacturing with predictive maintenance.

IoT Architecture and Protocols

• IoT architecture consists of four layers: perception, network, edge & cloud, and application.
• The perception layer captures data such as temperature, motion, and humidity using devices and sensors.
• The network layer transfers data via Wi-Fi, 5G, LoRaWAN, etc.
• The edge & cloud layer analyzes data on cloud or edge servers.
• The application layer provides user interaction through mobile apps, dashboards, and AI-driven automation.

Perception Layer

• The perception layer captures data from the physical world using sensors.
• Temperature sensors like DHT11 and DS18B20 are used in smart thermostats.
• Motion sensors (PIR sensors) are used in security systems to detect movement.
• Humidity sensors are used to monitor soil moisture levels in agriculture.

Network Layer

• The network layer transmits data from IoT devices to processing units.
• Wi-Fi is used for high-speed data transfer in smart home devices like Amazon Echo and Google Nest.
• 5G enables real-time monitoring in autonomous vehicles and smart cities with low latency.
• LoRaWAN is used in industrial IoT for long-range, low-power data transmission.

Edge & Cloud Layer

• The edge and cloud layer processes and stores collected data.
• Edge computing allows self-driving cars to process sensor data locally for quick decisions.
• Cloud storage allows IoT-enabled healthcare systems to store patient data on platforms like AWS and Azure for remote monitoring.

Application Layer

• The application layer provides interfaces for end-users to interact with IoT devices through mobile applications and automation software.
• Smartphone apps control IoT-enabled smart lighting like Philips Hue.
• AI-driven automation, such as Google Assistant and Amazon Alexa, enables voice-controlled smart home automation.

Security and Protocols

• The IOT-A protocol suite enables secure and efficient communication between IoT devices and the Internet.
• IoT devices connect to the Internet to collect and transmit data.
• Constrained domains are networks with limited resources.
• IoT gateways connect constrained domains to the Internet.
• In network scenarios, IoT devices communicate within a constrained domain or through an IoT Gateway.
• Protocol architecture includes IoT devices, gateways, and servers.
• Security procedures include authentication, encryption, and key establishment to secure communication.

End-to-End Secure Channel Establishment

• The goal is to establish secure communication channels between IoT devices and servers.
• Computationally heavy security functionalities are offloaded to IoT Gateways.
• An IoT device sends a request to an IoT Gateway, which establishes a secure connection with a server as an example.

Cooperative Security Protocols

• The goal is to enable IoT devices to securely communicate with each other.
• Cryptographic computational load is delegated to less resource-constrained nodes.
• An IoT device requests assistance from a more powerful node to establish a secure connection with another IoT device, serving as an example.

Key Terms Explained

• A constrained domain refers to a network that has limited resources available.
• An IoT Gateway acts as a device that connects constrained domains to the Internet.
• Cooperative Security is a strategy to delegates cryptographic computational load to less resource-constrained nodes.

IoT Network and Terminal Types

• Unconstrained Networks (NTU) are high-speed networks with low latency, for example, wired Internet.
• Constrained Networks (NTC) are low-speed networks with high latency, for example, IEEE 802.15.4.
• Unconstrained Terminals (TT1) are devices with sufficient power, such as internet servers and laptops.
• Constrained Terminals (TT2) have limited power, such as IoT devices.
• Tag-Type Terminals (TT3) have limited power, such as NFC tags.

Terminal and Network Constraints

• Power (P) classifications include battery-less (P0), battery-constrained (P1), green (P2), and power-unconstrained devices (P3).
• Code Size (E) classifications range from single loop controllers with MAC (E0) to multiprocess controllers with a full network IP stack (E5).
• RAM (R) classifications include stateless devices (R0) to devices capable of serving tens of packets and cached requests (R3).
• Computing Power (C) classifications includes computing power less than 8 MHz (C0) to around 100 MHz (C2).
• Network constraints include small frame sizes, low bitrates, high packet loss rates, delays, and disconnections.

IoT Communication Protocols

• MQTT is a lightweight protocol for low-bandwidth applications like remote monitoring.
• CoAP is suitable for low-power devices in industrial and smart city applications, like environmental monitoring.
• HTTP/HTTPS is used in web-based IoT systems requiring internet connectivity, like smart refrigerators.
• LoRaWAN is a low-power, long-range protocol for industrial and agricultural IoT.

Device Management

• Device registration and configuration ensures secure communication.
• Firmware updates fix vulnerabilities and improve performance.
• Security and access control prevents unauthorized access, using measures like multi-factor authentication.

IoT Devices and Sensors

• Microcontrollers (MCUs) are low-power computing devices used in IoT applications such as automating irrigation with soil moisture sensors and the ESP8266 (Wi-Fi enabled)

• System-on-Chip (SoC) integrates processor, memory, and connectivity, as seen in smartwatches processing health data using Qualcomm Snapdragon.

• FPGA (Field-Programmable Gate Array) uses customizable circuits for specialized IoT tasks, such as real-time processing in robotics.

• IoT devices use sensors to collect data.

• IoT devices use actuators to perform actions.

Sensors and Actuators

• Temperature sensors (e.g., DHT11) monitor climate conditions in smart homes.
• Motion sensors (e.g., PIR Sensor) are used in security systems.
• GPS modules enable location tracking in logistics.
• Actuators, such as motors, control robotic arms in industrial automation.
• Smart LED lights (e.g., Philips Hue) adjust brightness and color remotely via an app.

IoT Device Security

• Security threats include hacking, data breaches, and malware attacks.
• Secure Boot prevents unauthorized firmware modifications.
• Encrypted Communication uses AES and RSA encryption to secure data transmission.
• Firewall Protection blocks malicious access attempts.

IoT Data Characteristics

• IoT generates large-scale data (Volume) such as terabytes daily from smart cities.
• IoT requires real-time processing (Velocity) like in stock market predictions.
• IoT includes structured (databases), semi-structured (JSON), and unstructured data (Variety).

Data Analytics

• Descriptive Analytics explains past events with examples such as smart home energy usage reports.
• Predictive Analytics forecasts future trends, like car engine failure prediction based on sensor data.
• Prescriptive Analytics recommends actions, such as adjusting thermostat settings in smart homes.

Machine Learning in IoT

• Supervised Learning uses labeled data for security threat detection
• Unsupervised Learning detects anomalies in networks.
• Reinforcement Learning improves navigation from rewards, such as in self-driving cars.