An IoT system is made up of multiple layers that work together seamlessly. Understanding each layer helps you design, build, and troubleshoot IoT solutions. The four main layers are: Perception, Network, Processing, and Application.

The Four-Layer IoT Architecture

Layer 1: Perception Layer (Things Layer)

This is the physical layer ; the “eyes and hands” of the IoT system. It includes all the hardware that interacts with the real world.

• Sensors: Devices that collect data. Examples: temperature sensors (DHT11/DHT22), motion sensors (PIR), light sensors (LDR), humidity sensors, GPS modules, accelerometers.
• Actuators: Devices that perform actions. Examples: motors, relays, LEDs, buzzers, solenoid valves, servo motors.
• Microcontrollers/Processors: The “brain” of the device that controls sensors and actuators. Examples: Arduino Uno, ESP32, Raspberry Pi.
• Power Supply: Batteries, solar panels, or mains power that keep devices running.

Layer 2: Network Layer (Connectivity Layer)

Once data is collected, it must be transmitted. The network layer handles communication between devices and between devices and the cloud.

Technology	Characteristics	Common Use
Wi-Fi	High bandwidth, short range	Smart home devices, IP cameras
Bluetooth / BLE	Low power, very short range	Wearables, health monitors
Zigbee / Z-Wave	Low power, mesh network	Smart lighting, building automation
LoRa / LoRaWAN	Very long range, low power	Agriculture, smart cities
NB-IoT / LTE-M	Cellular, wide area coverage	Asset tracking, meters
Ethernet	Wired, high reliability	Industrial IoT, fixed installations

Layer 3: Processing Layer (Middleware / Fog / Edge Layer)

Raw data from sensors is processed at this layer. Processing can happen at three levels:

• On-device (Edge): The microcontroller processes data locally for fast responses. Ideal when low latency is critical (e.g., an alarm that triggers immediately).
• Fog Computing: Processing happens on a local gateway or edge server before sending to the cloud. Reduces bandwidth and latency.
• Cloud Computing: Data is sent to remote servers for heavy processing, storage, analytics, and machine learning. Examples: AWS IoT Core, Google Cloud IoT, Microsoft Azure IoT Hub.

Layer 4: Application Layer

This is where end users interact with the IoT system. The application layer presents processed data in a useful format and allows control of devices.

• Dashboards: Visual interfaces showing real-time sensor data, charts, and alerts (e.g., Grafana, ThingsBoard).
• Mobile Apps: Smartphone applications for monitoring and control.
• APIs: Interfaces that allow third-party software to access IoT data.
• Automation Rules: Logic that triggers actions based on conditions (e.g., “if temperature > 30°C, turn on fan”).

NOTE	In simple IoT projects, these four layers might run on just two devices: a microcontroller and a smartphone app. In industrial systems, each layer can involve hundreds of components.

Hardware Components in Detail

Microcontrollers vs Microprocessors

Primary role	Control sensors and actuators
Operating system	Usually none (bare-metal)
Power consumption	Very low (milliwatts)
Examples	Arduino Uno, ESP8266, ATmega328
Best for	Simple, dedicated IoT tasks

 

Common Sensors You Will Use in This Course

Sensor	Function	Application
DHT11 / DHT22	Temperature and humidity sensing	Weather stations, HVAC control
PIR Sensor	Detecting motion via infrared	Security systems, automatic lights
Ultrasonic (HC-SR04)	Measuring distance using sound	Obstacle avoidance, water level
LDR (Light Dependent Resistor)	Measuring light intensity	Automatic street lights
MQ-2 / MQ-135	Detecting gas concentrations	Air quality, smoke detectors
GPS Module (NEO-6M)	Determining geographic location	Vehicle tracking, mapping

Software Components

IoT systems require software at every layer:

 

• Firmware: Low-level code running on microcontrollers, written in C/C++ or MicroPython.
• Device drivers: Software that allows the processor to communicate with specific hardware.
• Communication middleware: Libraries and protocols that handle data transmission (e.g., MQTT client libraries).
• Cloud platforms: Software services for ingesting, storing, and visualizing IoT data.
• Analytics engines: Tools for processing large volumes of sensor data and identifying patterns.

Check Your Understanding — Lesson 0.2

1. Draw and label the four layers of the IoT architecture with one example component at each layer.
2. A farmer wants to monitor soil moisture remotely and automatically water crops. Name two sensors, one communication technology, and one type of actuator they would need.
3. What is the difference between edge computing and cloud computing? When would you choose one over the other?

Watch this Video