I2C SPI USB CAN eSPI Cable Testing View All Videos Quick Start Guides Software Downloads App Notes White Papers User Manuals Knowledge Base Sales Support About Us
Products Blog Sales Support Contact Account Search
A Simple Guide to IoT Architecture
Staff Writer

The Internet of Things (IoT) is the latest step in the evolution of connected information technology systems. A report published by Gartner predicts that the number of IoT connected devices deployed across the world will reach 20.4 billion by the year 2020 - that's nearly three times the total population of Earth. To put that into perspective, there were just 3.8 billion IoT devices deployed globally in 2015, and just 7 billion in 2018. As a growing number of consumers and corporations adopt IoT technology and solutions, the total number of internet-connected IoT devices is growing exponentially.

The IoT is a set of technologies that is both complex and elegant in its execution. We have the ability to build sensors that measure almost anything in the environment. If we connect those sensors to a smart devices with Wi-Fi and a microprocessor, we can collect readings from those sensors and send them to users or operators over the internet. With the right tools, we can turn that data into actionable information that a person (or machine) could use to make a decision or assess the health of a system or anticipate a problem.

Mobile remote control is where the "magic" of the IoT really happens. Instead of checking a sensor for a reading, readings are delivered to the user automatically. Sensors can also be connected to actuators, devices that allow things to do an action (switching a light or fan on or off, changing the speed of a motor, lowering the temperature in an environment, triggering an alarm, locking a door, etc.). Users can exercise remote control of these functions through smart devices that are connected to the IoT. 

IoT architecture refers to the set of underlying systems that are used to deliver services using the Internet of Things. Smart homes, automated warehouses, digital factories, and connected hospitals all rely on the same basic underlying infrastructure to deliver their critical IoT capabilities. If you are developing a product that will deliver services using the IoT, you can use this article to learn about IoT network architecture concepts and understand the basic components of IoT architecture.

iot architecture


The Three Layers of IoT Architecture 

IoT architecture can be described as a technology stack with three layers: the IoT device layer, the IoT gateway layer, and the IoT platform layer. Data originates in the device layer and passes through the gateway layer before entering the cloud where the IoT platform layer resides. Each layer plays an important role in the delivery of IoT services.

IoT Device Layer

The IoT device layer consists of all the smart devices that are connected to the system. Smart devices are products or assets that are embedded with sensors, processors, actuators, and the capability of transmitting data over the internet. They can collect data from their environment and share it with operators, users, other smart devices, and applications connected to the system. 

Smart devices can use many different types of sensors to collect data from their environment. An IoT device used in agriculture might include soil moisture sensors that measure the water content of the soil, humidity sensors that measure air moisture content, and temperature sensors that measure the atmospheric temperature. A smart home installation might include smoke sensors for detecting fire, touch and motion sensors for security, and a light sensor for automating lighting in the home. 

The simplest IoT implementations collect data from just one device, such as a lone home security camera. Other implementations may incorporate hundreds or even thousands of devices, requiring a more robust back-end infrastructure to manage the data volume and operations.

IoT Gateway Layer

The IoT gateway layer sits between the IoT device layer and the IoT platform layer. This layer consists of a physical device or software program that collects data from smart devices and transmits it to the cloud. The gateway layer offers two practical benefits to the IoT architecture: load management via data pre-processing and security.

Some smart devices are equipped with sensors that generate thousands or tens of thousands of data points every second. Consider a set of 12 networked high-definition security cameras, each one recording surveillance footage in 4K. If all of this data were directly uploaded to the cloud, there would be issues with bandwidth, response times, and network transmission costs, and the whole thing might be cost prohibitive. The gateway layer can consist of a dedicated software program that pre-processes data before sending it on to the cloud.

Gateway devices can also play a role in securing data transmission from smart devices. Features like tamper detection, encryption, and hardware random number generators can be implemented to prevent malicious attacks against IoT devices and secure data that is moving to the cloud.

IoT Platform Layer

Once data from the IoT is uploaded to the cloud, it can be processed by tools and applications in the IoT platform layer. The platform layer consists of Edge IT and cloud-based or physical data centers that play a role in data analytics, management, and archiving. Applications that provide data transforming, analytics, monitoring, and other functions and services exist in the IoT platform layer. The IoT platform layer also includes tools that visualize processed sensor data on user-facing devices. 

4 Stages of IoT Architecture 

The four-stage IoT architecture model offers a general framework for implementing a network of smart devices that collect data from the environment, pass that data to the cloud through internet gateways, use edge IT for basic analytics and pre-processing, and ultimately store data in data centers or in the cloud. 


Smart devices interact with their environment using sensors and actuators. Sensors capture data from the environment and relay it to data centers and the cloud via internet gateways and edge IT. Actuators are a kind of motor that can control or move a mechanical system. Actuators can be activated by operators using commands that originate in the cloud and are passed to smart devices through Internet gateways and edge IT. 

There are many types of sensors, including:

  • Accelerometers 
  • Color Sensors
  • Flow Sensors
  • Level Sensors
  • Light Sensors
  • GPS Sensors
  • Humidity Sensors
  • Proximity Sensors
  • Rain Sensor
  • Soil Moisture Sensors
  • Temperature Sensors
  • Tilt Sensors

Actuators in smart devices use low-power, high-efficiency electric motors. They may be plugged in or powered by battery.

Internet Gateways

The data captured by sensors begins its life cycle in analog form. Data from sensors must be aggregated, digitized, and transformed into a common format so it can be processed efficiently downstream. 

Internet gateway systems are typically located close to the sensors and actuators that generate the data. For example, a smart home system might include six HD cameras that each send data through a physical wire to an on-site data acquisition system (DAS). The DAS aggregates data from the sensor network, digitizes it, does some pre-processing, and may compress it to reduce its size before sending it down the pipeline. 

Edge IT 

Edge computing is all about facilitating computing transactions closer to the source to reduce latency and manage the load on data centers. Edge IT systems are the first stop for data that is uploaded from your sensors to the cloud. Here, specialized applications can be used to conduct analytics, parse large amounts of data for anomalies or KPI violations, and generate meaningful insights before passing them on to the data center. 

Data Center/Cloud

Data that requires deeper processing or analysis may ultimately be forwarded to a physical data center or to a cloud-based data storage server. Here, operators can implement cutting-edge technologies to analyze data, combining it with information from other sources to get a deeper understanding of how the system is behaving. 

Data analytics can be used to understand how sensor data varies over time and in relation to other variables. Machine learning technology can be used to automate and optimize actuators in response to sensory data or based on user patterns. There may also be some type of data visualization technology implemented that applies user business logic to data and presents information to users in an easily understandable format (tables, graphs, charts, etc.)


IoT architecture consists of many systems working together to facilitate two-way data transfer and communication between smart devices and the consumers and corporations that use them. 

As you design and develop your own IoT-enabled device, you'll need to plan ahead to ensure its compatibility and efficient functioning with other elements of the IoT architecture. The best way to verify that your IoT device works according to specifications is through rigorous programming, testing, and emulation with the Total Phase Promira Serial Platform. With our debugging tools, embedded engineers can streamline the development of IoT devices that fit seamlessly into existing IoT architecture models.

Request a Demo and we'll show you how you can use our development platform and debugging tools to make your project a success.