Unpacking the layers: Network Communications 101
Why you should care about how computers talk to each other
Happy December, everybody. As Cyber Adjacent emerges from the ashes of a hectic November, I’m excited to talk to you about the basics of network communications, i.e. how your devices chit-chat. Let’s begin with a fundamental networking communications concept: ‘network reference models’.
I know it might sound boring, but these network reference models are helpful frameworks for understanding computer conversations. They ensure interoperability and standardisation across all networks and their layers so no computer is left out.
If you’ve never thought about it, you may have felt the way computers chat with each other happened organically as systems developed. However, as you’re about to discover, it was not that straightforward, so let’s get into it.
What are network reference models?
One generally learns two types of network reference models: the OSI Model and the TCP/IP Model.
OSI Model
The Open Systems Interconnection (OSI) Model was developed in the 1980s by the International Organisation for Standardisation (ISO). In this model, network communications are split into seven layers:
1. Physical Layer: This refers to the actual physical connection and the transmission of raw data, such as electrical signals or light pulses, over cables or wireless.
2. Data Link Layer: This layer organises data into packets and ensures error-free delivery between directly connected devices.
3. Network Layer: This layer determines how data is sent across different networks to reach the correct destination, often using IP addresses.
4. Transport Layer: It ensures that data is delivered accurately and in the correct sequence, managing errors like error checking and retransmitting if necessary.
5. Session Layer: This layer establishes, maintains, and ends communication sessions between devices, like starting and stopping conversations.
6. Presentation Layer: This layer translates data into a format the application layer can understand, such as converting encrypted or compressed data.
7. Application Layer: The layer where users interact with the network through applications like web browsers or email, which handle the data exchange.
TCP/IP Model
The TCP/IP model emerged in the 1970s as a practical and flexible networking solution. It was developed by Defense Advanced Research Projects Agency (DARPA) researchers and engineers to address real-world communication needs. With the rise of the Internet, it quickly gained dominance.
1) Network Interface Layer: This foundation handles the physical connection and data exchange between devices on the same network, such as through cables or wireless signals.
2) Internet Layer: Responsible for routing data across different networks. It ensures that the data packets reach the correct destination using IP addresses.
3) Transport Layer: This layer ensures reliable communication by managing how data is sent and received between devices. It checks for errors, handles retransmissions, and ensures data arrives in the correct order.
4) Application Layer: This is where users interact with the network through software, like web browsers or email. It manages how data is presented and communicated between applications.
So, what are the differences between the two? Think of it this way: OSI is the classroom favourite, while TCP/IP is the workhorse out in the field.
First, there are the number of layers. The OSI model has seven layers, while TCP/IP simplifies things to four. Since fewer TCP/IP layers exist, they must handle more responsibility.
A key advantage of the OSI model, though, is its granularity. It divides specific tasks into separate layers, making explaining and visualising what’s happening under the hood easier.
Both models ultimately perform the same job. They take application data, pack it into a format the network can understand, and ensure it arrives at the right destination.
So, which one’s better? Honestly, neither – they serve different purposes. In the real world, you’ll hear “OSI” thrown around a lot, especially when people talk about “layers,” even if they mean TCP/IP. If someone mentions "Layer 3/4," you should clarify which model they’re referencing. For example, in OSI, that’s the Network and Transport Layers, but in TCP/IP, it’s more streamlined.
Knowing both is essential. OSI helps you understand the nitty-gritty of network communication, while TCP/IP is what you’ll see in action.
But don’t be mistaken; there was once beef between the two models and their supporters, known as the Protocol Wars.
The OSI was developed by industry representatives who wanted a multilayered system that allowed new collaboration possibilities across networks. Despite its comprehensive design, the OSI model struggled to gain traction due to the complexity and lengthy, consensus-driven standardisation process (literally thousands of engineers and policymakers were consulted) that couldn’t keep up with the pace of technology (luckily not an issue we have to face anymore, not.)
The TCP/IP model was more pragmatic and flexible, allowing for quick adaptation and widespread implementation. It emerged because of practical needs and was done by developers who needed something to effectively meet their emerging networking needs.
It’s a classic example of perfect being the enemy of good, but what also helped TCP/IP was the spread of the Internet, and the use of protocols that fall under the TCP/IP model also helped TCP/IP.
What else can I take from this?
Besides being something interesting to know about, and who doesn’t love hearing about historical beef, there is a practical benefit to someone who isn’t a networking expert learning about network reference models:
1) Troubleshooting: When your internet isn’t working, you can examine the layers to determine the cause. Is it physical, such as a cable not being plugged in or at a higher level, such as software issues?
2) Understanding how to protect yourself online: The OSI model explains how data moves and provides a good baseline for securing that data at various layers. For example, encryption is used at the presentation layer, VPNs at the network layer, and firewalls at the transport layer.
3) Understanding your threat environment: You can use the OSI model to explain where online threats happen, such as phishing (layer 7), malware (layers 7, 4 and 3) and rogue wireless access points (layers 1 and 2).
This is only the very basics of network communications. In 2025, we’ll dive deeper into it and discuss things like protocols. But that’s all for now. I hope you can have a relaxing end-of-year break and somehow incorporate your newfound knowledge of network communications into a conversation. If not, there is always next year!
Stay secure and eat some lasagna (or cake, onions, or anything with layers).