Architecture of IoT

(University of Michigan at Ann Arbor)

Every IoT system is different, and there are multiple ways of looking at IoT architecture. However, although there are many variations, they all follow the same basic structure and process. 

An IoT architecture requires four main components: connected hardware “things”, sensors and control devices that collect and process data via edge communication links (sometimes more than one), and transmit data to/from the cloud, a processing system (often part of a cloud platform offering), and cloud-based or local data centers for storing, managing, and analyzing data.

For IoT solutions to operate effectively, these four IoT architecture layers are required, each with specific functionality.

• Sensing Layer (or perception layer)
• Network Layer
• Data processing Layer
• Application Layer

The IoT device layer is the first layer of an IoT architecture. It consists of devices, sensors, and actuators that collect data from their surroundings and control things at the edge. These devices connect to the network layer through wired or wireless communication protocols. 

Devices that sense or control real-world things are the foundation layer of the IoT ecosystem. 

The IoT device layer includes the following elements: 

• Devices that sense or control things in the real world
• Sensors
• Cameras
• Actuators

The IoT device layer (or perception layer) is critical to allowing digital data to move between all elements of the IoT structure. It uses technologies such as HTTP, MQTT, and AMQP to facilitate transmission from one application/device to another.

The IoT network layer, also known as the transmission layer, acts as a bridge between the sensing layer and the application layer. The network layer of an IoT architecture is responsible for providing communication and connectivity between devices in the IoT system. It includes protocols and technologies that enable devices to connect and communicate with each other and with the wider internet. 

The IoT network layer has two sublayers:

• Routing layer: Handles the transfer of packets from source to destination
• Encapsulation layer: Another sublayer of the network layer

The IoT network layer carries and transmits information collected from physical objects through sensors. The transmission can be wireless or wire based. 

The IoT architecture has seven protocol layers: Application, Presentation, Session, Transport, Network, Data link, Physical. 

The IoT refers to the collective network of connected devices and the technology that facilitates communication between devices and the cloud, as well as between the devices themselves. Examples of network technologies that are commonly used in IoT include WiFi, Bluetooth, Zigbee, and cellular networks such as 4G and 5G. 

Additionally, the network layer may include gateways and routers that act as intermediaries between devices and the wider internet, and may also include security features such as encryption and authentication to protect against unauthorized access.

The data processing layer of IoT architecture refers to the software and hardware components that are responsible for collecting, analyzing, and interpreting data from IoT devices. The data processing layer receives raw data collected from sensors (or devices) and processes it into useful information. It accumulates, stores, and processes data that comes from the previous layer.

The data processing layer includes a variety of technologies and tools, such as data management systems, analytics platforms, and machine learning algorithms. These tools are used to extract meaningful insights from the data and make decisions based on that data.

The data processing layer consists of:

• Software and hardware components
• Cloud computing systems
• Big data modules
• Middleware layer

Example of a technology used in the data processing layer is a data lake, which is a centralized repository for storing raw data from IoT devices. 

The processing layer is also known as : the middleware layer, the brain of the IoT ecosystem.

The application layer of IoT architecture is the topmost layer that interacts directly with the end-user. It is responsible for providing user-friendly interfaces and functionalities that enable users to access and control IoT devices.

This layer includes various software and applications such as mobile apps, web portals, and other user interfaces that are designed to interact with the underlying IoT infrastructure. It also includes middleware services that allow different IoT devices and systems to communicate and share data seamlessly.

The IoT application layer also includes analytics and processing capabilities that allow data to be analyzed and transformed into meaningful insights. This can include machine learning algorithms, data visualization tools, and other advanced analytics capabilities. 

The IoT application layer acts as an interface between the user and IoT. The user can use the data and manage the device through the application that runs in this layer. The IoT layer can be customized to meet the requirements. 

Here are some IoT application layer protocols:

• Constrained Application Protocol (CoAP): A specialized web transfer protocol for use with constrained nodes and networks in the IoT. It runs over UDP and uses HTTP commands like GET, POST, PUT, and DELETE.
• Advanced Message Queuing Protocol (AMQP): An open standard for passing business messages between applications or organizations. It connects systems, feeds business processes with information, and transmits instructions to achieve goals.
• DDS: A middleware protocol for data-centric connectivity. It provides low-latency, data connectivity, extreme reliability, and a scalable architecture for business and mission-critical IoT applications.
• MQTT: A machine-to-machine (M2M) protocol used to communicate device data to servers. The main purpose of MQTT is to manage IoT devices.
• XMPP: Supports the real-time exchange of structured but extensible data between multiple entities on a network. It's most often used for consumer-oriented IoT deployments, such as smart appliances.
• LoRaWAN: A communication protocol that builds on LoRa to connect devices to a network. It's a type of LPWAN standard designed for IoT applications. It uses very little power, has good coverage, and works well indoors.

The application layer handles data formatting and presentation and serves as the bridge between what the IoT device is doing and the network handoff of the data it produces. Five of the top protocols and their features that matter most to IoT are: Constrained Application Protocol.

[More to come ...]