What is the OSI Model?

The Open Systems Interconnection (OSI) model is a conceptual framework that divides network communications functions into seven layers. Sending data over a network is complex because various hardware and software technologies must work cohesively across geographical and political boundaries. The OSI data model provides a universal language for computer networking, so diverse technologies can communicate using standard protocols or rules of communication. Every technology in a specific layer must provide certain capabilities and perform specific functions to be useful in networking. Technologies in the higher layers benefit from abstraction as they can use lower-level technologies without having to worry about underlying implementation details.

Why is the OSI model important?

The layers of the Open Systems Interconnection (OSI) model encapsulate every type of network communication across both software and hardware components. The model was designed to allow two standalone systems to communicate via standardised interfaces or protocols based on the current layer of operation.

The benefits of the OSI model are given next.

Shared understanding of complex systems

Engineers can use the OSI model to organize and model complex networked system architectures. They can separate the operating layer of each system component according to its main functionality. The ability to decompose a system into smaller, manageable parts via abstraction makes it easier for people to conceptualize it as a whole.

Faster research and development

With the OSI reference model, engineers can understand their work better. They know which technological layer (or layers) they’re developing for when they create new, networked systems that need to communicate with each other. Engineers can develop networked systems and take advantage of a series of repeatable processes and protocols. 

Flexible standardization

The OSI model does not specify the protocols to use between levels, but rather the tasks that protocols perform. It standardizes network communication development so people can rapidly understand, build, and decompose highly complex systems—all  without prior knowledge of the system. It also abstracts details, so engineers don’t require the understanding of every aspect of the model. In modern applications, the lower levels of networking and protocols are abstracted away to simplify system design and development. The following image shows how the OSI model is used in modern application development.

What are the seven layers of the OSI model?

The Open Systems Interconnection (OSI) model was developed by the International Organization for Standardization and others in the late 1970s. It was published in its first form in 1984 as ISO 7498, with the current version being ISO/IEC 7498-1:1994. The seven layers of the model are given next.

Physical layer

The physical layer refers to the physical communication medium and the technologies to transmit data across that medium. At its core, data communication is the transfer of digital and electronic signals through various physical channels like fiber-optic cables, copper cabling, and air. The physical layer includes standards for technologies and metrics closely related with the channels, such as Bluetooth, NFC, and data transmission speeds.

Data link layer

The data link layer refers to the technologies used to connect two machines across a network where the physical layer already exists. It manages data frames, which are digital signals encapsulated into data packets. Flow control and error control of data are often key focuses of the data link layer. Ethernet is an example of a standard at this level. The data link layer is often split into two sub-layers: the Media Access Control (MAC) layer and Logical Link Control (LLC) layer. 

Network layer

The network layer is concerned with concepts such as routing, forwarding, and addressing across a dispersed network or multiple connected networks of nodes or machines. The network layer may also manage flow control. Across the internet, the Internet Protocol v4 (IPv4) and IPv6 are used as the main network layer protocols.

Transport layer

The primary focus of the transport layer is to ensure that data packets arrive in the right order, without losses or errors, or can be seamlessly recovered if required. Flow control, along with error control, is often a focus at the transport layer. At this layer, commonly used protocols include the Transmission Control Protocol (TCP), a near-lossless connection-based protocol, and the User Datagram Protocol (UDP), a lossy connectionless protocol. TCP is commonly used where all data must be intact (e.g. file share), whereas UDP is used when retaining all packets is less critical (e.g. video streaming).

Session layer

The session layer is responsible for network coordination between two separate applications in a session. A session manages the beginning and ending of a one-to-one application connection and synchronization conflicts. Network File System (NFS) and Server Message Block (SMB) are commonly used protocols at the session layer.

Presentation layer

The presentation layer is primarily concerned with the syntax of the data itself for applications to send and consume. For example, Hypertext Markup Language (HTML), JavaScipt Object Notation (JSON), and Comma Separated Values (CSV) are all modeling languages to describe the structure of data at the presentation layer. 

Application layer

The application layer is concerned with the specific type of application itself and its standardized communication methods. For example, browsers can communicate using HyperText Transfer Protocol Secure (HTTPS), and HTTP and email clients can communicate using POP3 (Post Office Protocol version 3) and SMTP (Simple Mail Transfer Protocol).

Not all systems that use the OSI model implement every layer.

How does communication happen in the OSI model?

The layers in the Open Systems Interconnection (OSI) model are designed so that an application can communicate over a network with another application on a different device, no matter the complexity of the application and underlying systems. To do this, various standards and protocols are used to communicate with the layer above or below. Each of the layers is independent and only aware of the interfaces to communicate with the layer above and below it. 

By chaining together all these layers and protocols, complex data communications can be sent from one high-level application to another. The process works as follows:

  1. The sender’s application layer passes data communication down to the next lower layer.
  2. Each layer adds its own headers and addressing to the data before passing it on. 
  3. Data communication moves down the layers until it is eventually transmitted through the physical medium.
  4. At the other end of the medium, each layer processes the data according to the relevant headers at that level. 
  5. At the receiver end, data moves up the layer and is gradually unpacked until the application at the other end receives it.

What are alternatives to the OSI model?

Various networking models were used in the past, such as Sequenced Packet Exchange/Internet Packet Exchange (SPX/IPX) and Network Basic Input Output System (NetBIOS). Today, the main alternative to the Open Systems Interconnection (OSI) model is the TCP/IP model.

The TCP/IP model

The TCP/IP model is comprised of five different layers:

  • The physical layer
  • The data link layer
  • The network layer
  • The transport layer
  • They application layer

While layers like the physical layer, network layer, and application layer appear to map directly to the OSI model, this isn’t quite the case. Instead, the TCP/IP model most accurately maps to the structure and protocols of the internet.

The OSI model remains a popular networking model to describe how networking operates from a holistic perspective for educational purposes. However, the TCP/IP model is now more commonly used in practice.

A note on proprietary protocols and models

It’s important to note that not all internet-based systems and applications follow the TCP/IP model or the OSI model. Similarly, not all offline-based networked systems and applications use the OSI model or any other model.

Both the OSI and TCP/IP models are open standards. They’re designed so that anyone can use them, or further build them out to meet specific requirements.

Organizations also design their own internal, proprietary standards, including protocols and models, that are closed-source and only for use within their systems. Sometimes, they may subsequently release them to the public for interoperability and further community development. An example is s2n-tls, a TLS protocol that was originally a proprietary Amazon Web Services (AWS) protocol but is now open source.

How can AWS meet your computer networking requirements?

AWS helps organizations design, deploy, and scale networked systems and applications with less friction. 

We have a robust suite of AWS Networking and Content Delivery offerings. They’re designed to complement and integrate with your internal applications and services, across all levels of network operations. Here are some examples:

  • AWS App Mesh provides secure, application-level networking for all your services, with built-in communications monitoring and control
  • Amazon CloudFront is a content delivery network (CDN) service built for high performance, security, and developer convenience
  • AWS Direct Connect offers a direct connection, which doesn’t touch the internet, from your organization to your AWS resources
  • Elastic Load Balancing (ELB) distributes incoming network traffic across AWS targets to improve application scalability

Get started with networked systems and applications on AWS by creating an account today.

Next Steps on AWS

Sign up for a free account

Instant get access to the AWS Free Tier.

Sign up 
Start building in the console

Get started building in the AWS management console.

Sign in