The OSI model is a seven-layer conceptual framework that standardizes how network devices communicate by dividing the communication process into distinct functional layers. The International Organization for Standardization (ISO) published the OSI model in 1984 as standard ISO/IEC 7498 to allow products from different vendors to interoperate. This article defines the OSI model, explains each of the seven layers and its function, describes how data moves down the sending stack and up the receiving stack through encapsulation, compares the OSI model with the TCP/IP model, and explains why the OSI model matters for network troubleshooting.
The seven layers are the Physical, Data Link, Network, Transport, Session, Presentation, and Application layers, numbered one through seven from bottom to top. Each layer serves the layer above it and relies on the layer below it. The OSI model remains a reference structure rather than a protocol that runs on a wire, so understanding each layer clarifies where a specific network function or fault belongs.
What Is the OSI Model?
The OSI model is a seven-layer reference framework that describes how data passes between applications across a network, defined by the International Organization for Standardization. OSI stands for Open Systems Interconnection. The model separates network communication into seven layers, with each layer handling one defined part of the process and communicating only with the layers directly above and below it.
The OSI model is conceptual, which means no single product implements all seven layers as separate pieces of code. The framework instead gives engineers a shared vocabulary for the functions a network performs.
A function such as routing belongs to one specific layer, so naming the layer names the function. The OSI model sits alongside the practical TCP/IP protocol suite that actually carries traffic on the modern internet.
What Are the Seven Layers of the OSI Model?
The OSI model contains seven layers, numbered from Layer 1 at the bottom to Layer 7 at the top. The table below lists each layer, its primary function, and example protocols or units that operate at that level.
| Layer | Name | Function | Examples |
|---|---|---|---|
| 7 | Application | Provides network services to user programs | HTTP, FTP, SMTP, DNS |
| 6 | Presentation | Translates, encrypts, and compresses data | TLS, SSL, JPEG, ASCII |
| 5 | Session | Opens, manages, and closes sessions | NetBIOS, RPC, sockets |
| 4 | Transport | Delivers segments and controls reliability | TCP, UDP |
| 3 | Network | Routes packets between networks by IP address | IP, ICMP, routers |
| 2 | Data Link | Frames data and uses MAC addresses | Ethernet, switches, ARP |
| 1 | Physical | Transmits raw bits over the medium | Cables, hubs, radio, voltage |
The layers divide into two groups. Layers 1 through 4 handle data transport across the network, and layers 5 through 7 handle application and user-facing functions. A device such as a switch operates at Layer 2, and a router operates at Layer 3.
How Does the Physical Layer Work?
The Physical layer, Layer 1, transmits raw bits as electrical, optical, or radio signals over the physical medium. The Physical layer defines cables, connectors, pin layouts, voltage levels, and data rates. The Physical layer has no concept of addresses or data meaning, because the Physical layer moves only ones and zeros.
Ethernet cabling standards such as IEEE 802.3 and the Wi-Fi radio standards under IEEE 802.11 set the rules for the Physical layer. A hub repeats signals at the Physical layer without inspecting them. Common faults at the Physical layer include a disconnected cable, a damaged connector, or electrical interference on the line.
How Does the Data Link Layer Work?
The Data Link layer, Layer 2, packages bits into frames and delivers them between devices on the same local network using MAC addresses. The Data Link layer adds a header containing the source and destination hardware address. Each network interface card carries a unique MAC address that the Data Link layer uses to identify the sending and receiving device.
A switch forwards frames at Layer 2 by reading the destination MAC address. The Institute of Electrical and Electronics Engineers (IEEE) divides the Data Link layer into the Logical Link Control sublayer and the Media Access Control sublayer. Error detection through a frame check sequence also occurs at the Data Link layer.
How Does the Network Layer Work?
The Network layer, Layer 3, routes packets across separate networks using logical IP addresses. The Network layer determines the path a packet takes from the source network to the destination network. A router operates at the Network layer and forwards packets based on the destination IP address.
The Internet Protocol, defined by the Internet Engineering Task Force (IETF), is the dominant Network layer protocol. The Network layer also handles fragmentation, which splits a packet to fit the maximum size a link allows. The Internet Control Message Protocol (ICMP) carries error and diagnostic messages at the Network layer, and the ping command relies on ICMP.
How Do the Transport, Session, Presentation, and Application Layers Work?
The upper four layers handle reliable delivery and user-facing services. Each layer below performs one defined role.
- Transport layer (Layer 4) controls reliability and flow. The Transport layer segments data and uses TCP for ordered, acknowledged delivery or UDP for fast, connectionless delivery.
- Session layer (Layer 5) manages conversations. The Session layer opens, maintains, and closes the dialogue between two applications and can restore a session after an interruption.
- Presentation layer (Layer 6) translates data formats. The Presentation layer encrypts, compresses, and converts data so the Application layer receives it in a usable form, including TLS encryption.
- Application layer (Layer 7) delivers network services. The Application layer provides protocols such as HTTP, SMTP, and DNS that user programs call directly.
The Transport layer divides into TCP and UDP. TCP guarantees delivery through acknowledgments and retransmission, and UDP omits those checks to reduce delay for traffic such as video and voice.
How Does Data Move Through the OSI Layers?
Data moves down the seven layers on the sending device and up the seven layers on the receiving device through a process called encapsulation. Each layer adds its own header to the data it receives from the layer above before passing it down.
- The Application layer generates the data from a user program such as a web browser.
- The Presentation and Session layers format the data and establish the session.
- The Transport layer splits the data into segments and adds port numbers for TCP or UDP.
- The Network layer wraps each segment in a packet and adds source and destination IP addresses.
- The Data Link layer wraps each packet in a frame and adds source and destination MAC addresses.
- The Physical layer converts the frame into bits and transmits the signal over the medium.
The receiving device reverses the process. Each layer reads and removes its matching header in a step called decapsulation, then passes the remaining data up to the next layer until the Application layer delivers it to the program.
How Does the OSI Model Compare to the TCP/IP Model?
The OSI model uses seven layers as a teaching reference, while the TCP/IP model uses four layers that describe how the internet actually operates. The TCP/IP model combines several OSI layers into single layers.
| OSI Layer | TCP/IP Layer | Role |
|---|---|---|
| Application, Presentation, Session | Application | User services, formatting, sessions |
| Transport | Transport | End-to-end delivery with TCP or UDP |
| Network | Internet | Routing packets with IP |
| Data Link, Physical | Network Access | Framing and physical transmission |
The TCP/IP model came from the United States Department of Defense ARPANET work and predates the OSI standard. The OSI model offers finer separation for diagnosis, and the TCP/IP model describes the protocols that run on the public internet. The full structure of the practical stack appears in the overview of the TCP/IP model.
Why Does the OSI Model Matter for Troubleshooting?
The OSI model matters for troubleshooting because it lets a network engineer isolate a fault to one specific layer and rule out the others. A structured approach moves from Layer 1 upward or from Layer 7 downward.
- Check Layer 1 first for physical faults. A dead link, unplugged cable, or disabled port shows the problem sits at the Physical layer.
- Check Layer 2 for local delivery faults. A wrong MAC table entry or a VLAN misconfiguration points to the Data Link layer.
- Check Layer 3 for routing faults. A failed ping or wrong gateway points to an IP addressing or routing fault at the Network layer.
- Check Layers 4 through 7 for service faults. A blocked port, expired certificate, or DNS failure points to the Transport or Application layers.
Key Takeaways
- The OSI model defines seven layers. The Physical, Data Link, Network, Transport, Session, Presentation, and Application layers each handle one function.
- The ISO published the OSI model. The International Organization for Standardization released the framework in 1984 as ISO/IEC 7498.
- Encapsulation moves data down and up. Each layer adds a header on the way down and removes it on the way up.
- The OSI model has seven layers and TCP/IP has four. The TCP/IP model merges several OSI layers into single layers.
- The OSI model guides troubleshooting. Isolating a fault to one layer narrows the search for a network problem.
What are the 7 layers of the OSI model?
The seven layers are Physical, Data Link, Network, Transport, Session, Presentation, and Application, numbered one through seven from bottom to top. Each layer performs one defined network function.
Who created the OSI model?
The International Organization for Standardization (ISO) created the OSI model and published it in 1984 as standard ISO/IEC 7498. The model standardizes network communication across vendors.
Is the OSI model still used today?
The OSI model is used today as a teaching and troubleshooting reference. The internet runs on the TCP/IP model, but the OSI layers remain the standard vocabulary for describing network functions.
What is the difference between OSI and TCP/IP?
The OSI model uses seven layers as a conceptual reference. The TCP/IP model uses four layers that describe the protocols running on the actual internet. TCP/IP merges several OSI layers.
Which OSI layer does a router operate on?
A router operates at Layer 3, the Network layer, where it forwards packets between networks using IP addresses. A switch operates at Layer 2, the Data Link layer, using MAC addresses.
What is encapsulation in the OSI model?
Encapsulation is the process where each OSI layer adds its own header to the data from the layer above as the data moves down the stack before transmission across the network.
Last Thoughts on the OSI Model
The OSI model divides network communication into seven layers that move from raw bits at the Physical layer to user services at the Application layer. The International Organization for Standardization defined the framework so devices from different vendors interoperate and so engineers share one vocabulary for network functions. Each layer adds a header during encapsulation, and the receiving device removes each header in turn.
The model differs from the four-layer TCP/IP model that runs the public internet, yet the layered structure still guides fault isolation during troubleshooting. The addresses used at Layer 2 are explained in the overview of MAC addresses, the routing addresses at Layer 3 in the guide to IP addresses, and the practical stack in the overview of TCP/IP. The full set of networking topics sits on the how networks work hub.
