The seamless internet experience that we get today comes from years of experimentation and trial and error. The internet has come a long way since its inception, and there are two main models on which the internet works — The Open Systems Interconnections (OSI) model and the Transmission Control Protocol/Internet Protocol (TCP/IP) model.

The TCP/IP protocol puts a unique identifiable address on every device on the internet and hence makes it super easy and quick to communicate between devices. Notice that IP address your computer was assigned? That address makes a huge difference.

Also read: 5 ways to access blocked websites

What is TCP/IP? How does it work?

The TCP/IP is just a condensed, more refined version of the OSI model. It has four layers as compared to the OSI’s seven.

Which layers comes first though, depends upon whether you’re sending data or receiving it. If you’re sending, that’ll be the Process layer and if you’re receiving it’ll be the Network Access layer.

In this article, we’ll be talking about the client end.

Network Access Layer

This layer is the equivalent of the Data Link layer and Physical layer from the OSI model. This layer takes care of the hardware addressing the is primarily responsible for the physical transmission of data.

All the hardware concerned with networking is summed up in this layer, and here onwards, we go up.

Internet Layer

This is the equivalent of the Network layer from the OSI model. This layer defines all protocols that are needed for the logical transmission of data over a specific network, or over the internet.

It has the following protocols

• Internet Protocol: It looks at IP addresses assigned to individual devices and then delivers data packets from the source host to the destination host. It has two versions — IPv4 and IPv6, the latter being the more secure one.
• ICMP: The Internet Control Message Protocol resides inside of IP and makes sure that a host gets enough information about network issues.
• ARP: The Address Resolution Protocol takes on the job of finding physical address from a known host’s IP address. There’s also the reverse ARP, which you guessed it, finds the IP address from a known hardware address.

Also read: How to check the IP Address in Linux?

Host-to-Host Layer

The equivalent of the Transport layer from the OSI model, the Host-to-Host layer is perhaps the most notable of all layers in the TCP/IP model. This layer takes care of two most important things — error-free data delivery and end-to-end communication.

It has two main protocols:

• Transmission Control Protocol(TCP): The primary protocol here, of course, best known for providing reliable and secure communication between end systems. It has a bunch of security features like segmentation. However, all that extra work adds a lot of overhead and hence, more cost.
• User Datagram Protocol(UDP): This is almost the same as TCP, however without the reliability features. So while you’re losing out on reliability and security, you’re gaining cost-effectiveness and some speed. It is also worth noting that UDP is a connectionless protocol as compared to TCP, which is connection-oriented.

Process Layer

The final layer for the TCP/IP protocol takes on three different layers from the OSI model, namely — Application, Presentation and Session Layers.

This is the final layer on the receiving end and is primarily responsible for node-to-node communication and user-interface applications. All major communication/access protocols like HTTP/HTTPS, SSH, FTP/TFTP, Telnet, DNS, DHCP are located in this layer.

Also read: What is no-verity-opt-encrypt? Why is it used?