Understanding the IoT Protocol Stack: From Hardware to Application Layer

Understanding the IoT Protocol Stack: From Hardware to Application Layer

Some people call the Internet of Things (IoT) a big web of connected "things," but the real magic is in how they communicate with one other. The IoT Protocol Stack is a complex, multi-layered technology that allows data travel hundreds of miles from a vineyard's moisture sensor to a farmer's smartphone. The IoT stack is better than the old OSI model for many forms of communication, low power use, and efficiency. This is a step-by-step guide of how data gets from the real world to the cloud.

1.     The Physical and Data Link Layers: The Foundation At the absolute bottom of the stack are the Physical (PHY) and Data Link (MAC) layers. This is where the "thing" and the "internet" meet. This layer shows you how to change data into electrical, radio, or optical signals. When it comes to the Internet of Things (IoT), you often have to choose between range, power, and bandwidth:

• Short Range: Bluetooth Low Energy (BLE) and Zigbee are two technologies that are quite popular for smart homes and wearables since they don't use much energy.

• large Range: LoRaWAN and NB-IoT let sensors send small amounts of data over large distances. This makes them perfect for smart cities or farming.

• High Bandwidth: Wi-Fi and Ethernet are fast enough for things like security cameras that utilize a lot of data, but they take a lot of electricity.

2. The Network Layer is the Post Office. The Adaptation Layer of the Network Layer makes sure that data is transmitted to the proper place so that it can go to the internet whenever the physical hardware provides a signal. The biggest challenge is that most IoT devices are too small to use normal IPv6 packets. To fix the problem, developers use 6LoWPAN (IPv6 over Low-Power Wireless Personal Area Networks). This protocol "compresses" the big IPv6 headers so that they can fit into the little data packets that Zigbee and other low-power radios utilize. This layer gives each sensor its own "home address" on the internet.

 3. The Transport Layer (The Mailman) Sending data from one end to the other is the job of the Transport Layer. It makes sure that the server gets the data it obtains from the device. There are two main people who are involved:

• TCP (Transmission Control Protocol): This is the delivery service that you can trust. It checks to see if all the packets have arrived and asks for a resend if one is missing. It is used by protocols like MQTT.

• UDP, or User Datagram Protocol, is the service that sends things quickly. It sends data right away, without waiting for a response. It is less reliable, but it is much faster and uses less battery, which makes it perfect for sensors that provide data often (if one update is lost, it won't be a major concern).

 4. The Language Layer (Also known as the Messaging Layer) This is where the IoT stack is most different from the web. Even if your computer uses HTTP to browse the web, it's too "heavy" for a little sensor. On the other hand, IoT uses certain message protocols:

• MQTT (Message Queuing Telemetry Transport): The best way to link IoT devices. It works on the "Publish/Subscribe" model. A sensor sends out a temperature reading, and any app that has permission can obtain it. It's light and works well with connections that aren't always solid.

 • CoAP (Constrained Application Protocol): This is a custom protocol that looks like HTTP but works over UDP. They utilize it when devices need to talk to each other in a web-like way but don't have a lot of memory or power.

5. The Application Layer (The End Point) The data is useful at the top of the stack. This is the part of the application that the user interacts with. The raw bits from the sensor have now been packaged into a package, given an address, and delivered to a cloud platform like AWS IoT, Azure, or a private server. This is where the data is decrypted, looked at, and shown in an image. This is where "logic" takes place. For instance, if the moisture sensor reads less than 20%, the application layer sends a directive back down the stack to open an irrigation valve.

SR University School of ComputerScience, Artificial Intelligence , SR University