Imagine a world, where every mobile phone manufacturer uses a different charging port, a different voltage level, and a different charging method.
A charger from one company would NOT work with a phone from another company. Life would become extremely difficult.
Now imagine the same situation in computer networking.
In the early days of computer networks, different companies built their own networking systems. A computer manufactured by one company often could not communicate with a computer manufactured by another company.
Every manufacturer like IBM, Xerox, IBM, DEC and many others followed its own rules, communication methods, data formats, and hardware standards.
This problem became known as interoperability failure.
“Interoperability means is the ability of different systems to communicate and work together.”
.
As a result, networks were difficult to build and devices from different manufacturers could not communicate properly.
Scientists quickly realized that if computer networking was going to grow globally, there had to be a common framework that everyone could follow.
This need led to the development of networking models.
.
Networking models
provide a structured way to understand and design communication between devices. Instead of treating communication as one huge complicated process, networking models divide communication into smaller tasks called layers.
Today, almost every communication system in the world is based on two important networking models whether you are browsing a website, sending an email, watching a YouTube video, making a video call or anything related to networks.
OSI Model
(Open Systems Interconnection Model) developed by International Organization for Standardization (ISO) in 1984.
The OSI Model contains seven layers.
- Application Layer
- Presentation Layer
- Session Layer
- Transport Layer
- Network Layer
- Data Link Layer
- Physical Layer
The OSI model is primarily a reference model.
It helps students and engineers understand communication in a structured way.
.
TCP/IP Model
(Transmission Control Protocol / Internet Protocol Model) developed by the United States Department of Defense in January 1, 1983.
The TCP/IP model contains four layers.
- Application Layer
- Transport Layer
- Internet Layer
- Network Access Layer
This model is used in real Internet communication. Every website, email service, cloud platform, and mobile application relies on TCP/IP concepts.
.
Why Do We Need Networking Models?
Networking models solve several critical problems such as –
- Standardization – Manufacturers around the world can design products that follow the same rules. As a result, devices become compatible. Communication becomes reliable.
- Interoperability – different devices can communicate successfully. For example – Windows Laptop, Linux Server, Android Phone, Cisco Router, Cloud Server.
- Easier Troubleshooting – Suppose a user cannot access a website.
The problem could be damaged cable, Incorrect IP address, DNS issue, Application issue. Networking models allow engineers to identify the exact layer where the problem exists. Instead of checking everything, troubleshooting becomes systematic. - Modular Design – Each layer performs a specific responsibility. A change in one layer usually does not affect other layers.
- Faster Technology Growth – Because responsibilities are separated, engineers can improve one layer without redesigning the entire network.
For example – Wi-Fi technology improved dramatically over time. However, applications such as email and web browsing continued working because only lower-layer technologies changed.
.
Birth of Layering Models
Researchers realized that communication could be simplified if it were divided into smaller layers. Instead of creating one huge networking system, engineers could separate communication into independent tasks.
For example – One layer could handle physical transmission.
Second layer could handle addressing.
Third layer could handle error detection.
Fourth layer could handle application communication.
This layered approach made networking easier to understand, develop, and troubleshoot.
Good networking models exhibit several characteristics.
.
Characteristics of Layered Models
Separation of Responsibilities – Each layer performs a dedicated task.
Layer Independence – Changes in one layer have minimal impact on others.
Service-Oriented Design – Each layer provides services to the layer above it.
Encapsulation Support – Each layer can add its own control information.
Scalability – New technologies can be integrated without redesigning the entire system.
.
Layered Communication Concept
Consider two computers communicating.
Each layer communicates logically with its corresponding layer.
Transport Layer talks logically to Transport Layer.
Network Layer talks logically to Network Layer.
In reality, data travels downward through layers on the sender side and upward through layers on the receiver side. This concept is called peer-to-peer communication.
.
Advantages of Networking Models
Simplicity – Complex communication becomes easier to understand.
Standardization – Manufacturers follow common rules.
Interoperability – Devices from different vendors can communicate.
Easier Troubleshooting – Problems can be isolated layer by layer.
Flexibility – Technologies can evolve independently.
Better Maintenance – Changes can be implemented efficiently.
.
Disadvantages of Networking Models
Additional Overhead – Each layer adds headers and control information. This increases transmission size.
Processing Delay – Every layer performs additional processing.
Complexity for Beginners – Students often find multiple layers confusing initially.
Theoretical Separation – Some layers overlap in practical implementations.
This is especially true in the OSI model.
.
“Despite these limitations, networking models remain indispensable for understanding communication systems.”
Test Yourself
Q1- (IIT Bombay, B.Tech End Semester)
What is meant by interoperability in computer networks? How did the absence of a standard networking model affect early computer networks? How did networking models solve this problem?
Ans – Interoperability means the ability of different systems, devices, and software from different manufacturers to communicate and work together successfully over a network. A Windows laptop connecting to a Linux server, or a Cisco router forwarding packets from an Android phone — all of this is interoperability in action.
In the early days of networking, interoperability did not exist. Companies like IBM, DEC (Digital Equipment Corporation), and Xerox each built their own proprietary networking systems. IBM computers followed IBM’s rules. DEC computers followed DEC’s rules. The two could not communicate. If a university purchased IBM machines and its partner institution had DEC machines, data exchange required expensive, custom-built conversion hardware or physical transport of storage media. This problem was called interoperability failure and it severely limited the growth of networking.
Networking models solved this in two ways.
First, they established standardization — a common framework that all manufacturers could follow when designing products. If every manufacturer follows the same layer boundaries and interface rules, their devices can communicate regardless of internal implementation differences.
Second, they enforced separation of responsibilities through layering. Each layer has a defined job and a defined interface to the layers above and below it. As long as a manufacturer’s implementation respects these interfaces, it does not matter what technology is used internally.
Q2- (NIT Tiruchirappalli, B.Tech) — Explain the concept of layered communication with a suitable real-world analogy. What is meant by peer-to-peer communication in the context of networking models?
Ans – Layered communication means dividing the complex process of network communication into a series of smaller, manageable tasks, where each task is assigned to a specific layer. Each layer performs its own job independently and interacts only with the layers directly above and below it.
A very effective real-world analogy is the postal system. When you write a letter and post it, the process naturally divides into layers. You write the message — this is the Application Layer creating data. You seal it in an envelope — this is encapsulation, where the layer adds its own packaging. You write the addresses on the envelope — this is the addressing function of the Network Layer. The post office transports the envelope through sorting centres and delivery routes — this is routing and physical transmission. At the destination, the receiver opens the envelope and reads the message — this is decapsulation and delivery to the application. Crucially, you — the sender — never needed to understand how the post office sorted the mail internally. Each “layer” of the postal process did its own job.
Peer-to-peer communication is the concept that each layer on the sender’s side communicates logically with the corresponding layer on the receiver’s side, even though physically the data travels down through layers at the sender and up through layers at the receiver.
Q3- (IIT Madras, M.Tech Entrance) — What are the key characteristics of a well-designed layered networking model? Explain each characteristic with its significance in network design.
Ans – A well-designed layered networking model exhibits five key characteristics:
Separation of Responsibilities – This separation means that an engineer designing a routing protocol does not need to worry about whether the underlying medium is copper or fibre.
Layer Independence – A change in one layer should have minimal or no impact on other layers, as long as the interface between layers remains unchanged. The most powerful real-world example of this is Wi-Fi.
Service-Oriented Design — Each layer provides a service to the layer above it and uses the services of the layer below it.
Encapsulation Support — Each layer must be able to add its own control information (header, and sometimes trailer) to the data it receives from the layer above, before passing it downward.
Scalability — New technologies should be integrable into the model without requiring the entire system to be redesigned.
Q4- (Anna University, B.Tech Examination) – Compare the advantages and disadvantages of using a layered networking model. Is the overhead introduced by layering justifiable? Give reasons.
Ans –
Advantages
Simplicity – Complex communication becomes easier to understand.
Standardization – Manufacturers follow common rules.
Interoperability – Devices from different vendors can communicate.
Easier Troubleshooting – Problems can be isolated layer by layer.
Flexibility – Technologies can evolve independently.
Better Maintenance – Changes can be implemented efficiently.
Disadvantages
Additional Overhead – Each layer adds headers and control information. This increases transmission size.
Processing Delay – Every layer performs additional processing.
Complexity for Beginners – Students often find multiple layers confusing initially.
Theoretical Separation – Some layers overlap in practical implementations.
Is the overhead justifiable? Yes, overwhelmingly so. The alternative — monolithic, vendor-specific, non-interoperable networking — is far more costly in practical terms. The Internet serves over five billion users precisely because TCP/IP’s layered model allowed any device from any manufacturer to participate without custom engineering. The overhead of a few dozen bytes of headers is trivial compared to the alternative cost of custom conversion hardware at every network boundary. The productivity gained through global interoperability dwarfs any efficiency lost to header overhead.
Q5- Why is the OSI model called a reference model?
Ans – The OSI model is called a reference model because it was designed as a conceptual framework — a set of guidelines defining what each layer should do — rather than a set of specific protocols defining how to do it. No real network is literally built as seven distinct software modules labeled exactly as OSI describes. Instead, real-world protocols are designed to be broadly consistent with OSI’s layer responsibilities, and engineers use OSI as a common vocabulary for discussion and troubleshooting.
Q6- The OSI model has seven layers, but real networks do not implement all seven as separate modules. Does this make the OSI model useless? Justify your answer.
Ans – Not at all and this question reveals a fundamental distinction between a model and an implementation. The OSI model was never intended to be a blueprint for how to write networking software. It was intended as a conceptual framework, a way of thinking about the problem of network communication.
Q7- The OSI model was developed by which organization and in which year?
IEEE, 1980
ISO, 1984
IETF, 1983
DARPA, 1969
Ans – (2)
Explanation – The OSI (Open Systems Interconnection) model was developed by ISO, the International Organization for Standardization and published in 1984.
IEEE develops hardware-level standards like Ethernet (802.3) and Wi-Fi (802.11).
IETF develops Internet standards published as RFCs.
DARPA developed the TCP/IP protocol suite.
Q8- The concept that the Transport Layer at the sender communicates logically with the Transport Layer at the receiver, even though data physically travels down and up through all layers is called
Multiplexing
Encapsulation
Peer-to-peer communication
Synchronisation
Ans – (3)
Explanation – Peer-to-peer communication in the context of networking models means that each layer on the sending side communicates logically and directly with its corresponding layer on the receiving side.
Q9- Which of the following is a disadvantage of using a layered networking model?
Devices from different vendors cannot communicate
Each layer adds header information, increasing overall data size
Troubleshooting becomes more complex
Technologies at one layer cannot evolve independently
Ans – (2)
Explanation – each layer adds its own header (and sometimes a trailer in data link layer) to the data before passing it to the next layer. This increases the total size of the transmission compared to the original data.
