Introduction

Now we explain TCP/IP protocol suite, which we also call Internet protocol suite. In this blog, we will also discuss TCP/IP versus OSI model. All the points are useful in your competitive exams, even in your college university exams, interviews—anywhere you are asked, you can easily answer. So, quickly like this blog and subscribe to the channel if you have not done so till now. If you have done it, then get it subscribed from other devices. I am giving you a task: whoever is reading this blog, you have to bring at least one subscriber today itself. When you have done it, comment as done to acknowledge me.

TCP/IP vs. OSI Model

Overview

Let’s start. If we talk about TCP/IP versus OSI, I have already explained to you very well about OSI. There is a seven-layer architecture: physical layer, data link layer, network, transport, session, presentation, and application. But if we talk about TCP/IP protocol suite, then we have both 4-layer and 5-layer architectures. First, let’s see what is the difference between the 4-layer and 5-layer architectures.

4-Layer and 5-Layer Architecture

In 4-layer architecture, we have 4 layers. Here, physical layer and data link layer are combined to form network access layer. Network layer is called Internet layer. Transport layer is called transport layer or host-to-host layer. These three layers are combined and called application.

What are the 5 layers in actual?: Physical layer and data link layer are separate, and the others are the same. Now, why is there a difference? The major reason for the difference is that OSI was actually a pure theoretical model. Both of these are fundamental models. Fundamental means if we talk about the Internet, how communication works in the Internet, then always we talk about TCP/IP model. But this OSI model is a fundamental model in every communication. You can say if you want to read the network very well, then you should have good knowledge about this, but this whole is a theory. TCP/IP is a complete implementable model.

Who developed it?:  ARPANET developed it and was funded by DARPA, which was the defense agency of America. They funded it, and ARPANET actually implemented it for use. So this is actually an implementable model, and there is no documentation type of work. This was full theory, and this is full practical. So when we work in a practical approach, we don’t work on theory or documentation. Therefore, they didn’t work on this.

4-Layer vs. 5-Layer in Practice

So, in many books and in many materials, you will get 5 layers, and in many, you will get 4 layers. In actual, 4 layers are more important, and we should have good knowledge about it. But this depends on your college or university if they have books of frozen then this one, if they have Tanenbaum then this one. So this depends on what book you are reading. But the story of both is the same.

Application Layer

If we talk about the application layer, obviously, what is the major work of the application layer? Process-to-process delivery. Process-to-process delivery was here also, and it is here also. Along with this, the work of the presentation layer, let’s say if I am using a different encoding system, then to synchronize them, encryption, decryption—all that work now comes under this. Other than this, to generate data, whose work is it? Where is the user? Here means he is using some application, either SMTP if I want to send mail, FTP if I want to transfer a file, telnet for remote login. All applications, all the protocols are the same as they were in OSI.

Transport Layer

Then if we talk about the transport layer, what is its use? Host-to-host delivery. TCP model, which is the major because what is its name? TCP/IP protocol means Transmission Control Protocol is used and Internet Protocol (IP) network protocol we are using here. In which there is IPv4 and IPv6 also. So that came in your Internet layer.

In transport layer, you have majorly TCP, UDP, and STTP. So you want to make connectionless or connection-oriented, that you know, so that is your host-to-host. Then in source-to-destination, where we talk about IP addressing, adding IP address in header of source, destination, that is taken care of by Internet layer.

Network Access Layer

Then finally, we have physical or data link layer. Their responsibility is that within your own network, within one node to second node, how data is to be sent? That work is done by network access layer. So here is the diagram, the same protocol stack architecture. The user is here. In this way, it will go and one by one header will be added and go to the destination from here. This is stack architecture, so the same last-in-first-out case is there as it is in OSI.

Data Transmission Process

Now, you can say the user generated some data. He went to application layer. From application layer to transport, then from transport went to Internet layer, and from there he went to network access layer. So in network access layer, let’s say I have a node or a router. On the router, which two layers will be there? One, the network access layer, and the other one Internet layer.

In this way, the packet will move and go to the next one. And what will happen in the next? Same network access layer and Internet layer. So again, the packet will move, and when the receiver comes, then again it will move through these four layers. So the same architecture. In OSI, we talked about physical layer, data link layer, and network layer, so here within these two main layers, the whole story moves.

Main Protocols and Models

So these main protocols like IPv4, IPv6 came under this. IGMP, ICMP all came under this. If we talk about MAC protocols, error control, flow control within the local network, so that comes under the network. And if we talk about transport layer, I have already told you about TCP, UDP, and all that. So data term we use here, segment here, packet here, and frame here, which we were using as it is in OSI.

Implementable Model and Architecture

The main difference you have already figured out. Whenever we talk about an implementable model in the Internet, which model do we use? So TCP/IP model we use. Developed—you know this has been developed by ISO, and this was developed by ARPANET. And if we talk other than this, it supports client-server architecture and peer-to-peer.

Client-Server vs. Peer-to-Peer

What is meant by client-server?:  A centralized authority is there, server, and all clients are attached to it, and they send their request-response. And meaning of peer-to-peer is there is no centralized authority. All are mutually connected, and data is transferred in a distributed way. So both are supported.