What is the OSI model?
In order to ensure error-free communication between devices from different manufacturers and also to standardize it, a reference model has become established in IT: the ISO-OSI layer model. It divides the internet into seven different, finely structured layers, which all cooperate with and among each other and each take over certain tasks. This ensures systematic data transmission.
- A definition of the OSI model ➔
- How is the OSI model structured? ➔
- What are the layers in the OSI model? ➔
- How does the OSI model work in practice? ➔
- What are the advantages of the OSI model? ➔
- The disadvantages of the OSI model ➔
- What other layers are there? ➔
- The OSI model vs. the TCP/IP model ➔
- The OSI model: what you need to know ➔
A definition of the OSI model
What good is it if the other person speaks and uses a foreign language perfectly and without an accent, but you yourself do not have the necessary “tools” to understand it? This results in complete confusion. In order to successfully solve just this problem in the world of information technology, the OSI reference model has been used in computer and network technology for some time now. It partitions complex processes into individual steps (actually: layers). Each of these layers is assigned a specific task; they are superimposed on each other, guaranteeing that communication between PCs and similar devices is manufacturer-independent and cross-vendor. The model is very often used as a reference when communication or message transmission processes are depicted.
But what does the abbreviation ISO-OSI actually stand for? This is quickly explained: ISO is short for International Organization for Standardization, while OSI stands for Open Systems Interconnection, i.e. an “open” system for communication connections and networking.
OSI was developed starting around 1983 by representatives of large computer and telecommunications companies and was originally intended to be a detailed specification of the actual interfaces. Instead, the committee decided to create a common reference model that could then be used by other developers to design detailed interfaces that in turn could be used to create general purpose systems for transmitting data packets. The OSI architecture was officially adopted as an international standard by the International Organization for Standardization in 1984—a great success for the developers of the concept. The 1984 standard was eventually replaced by ISO/IEC 7498-1:1994. Other standards relating to the OSI model include DIN ISO 7498 and ITU-T X.200.
How is the OSI model structured?
As for how the model works, it is used by IT professionals worldwide to model or track how data is sent or received over a network. As mentioned above, the OSI model partitions data transmission into seven layers, each of which is responsible for performing explicitly defined tasks related to the sending and receiving of specific data. The concept of OSI is that the communication process between communicator and recipient in a network can be divided into seven different groups of related functions. Each communicating user or program is on a device that provides these seven functional layers—at the other end of communication is the appropriate counterpart: a receiving device that can read the information. Each of the seven layers serves the layer above it and is in turn served by the layer below it. In a particular message between users, data flows through the layers in the source computer, over the network, and then up through the layers in the receiving computer. Only the application layer “at the very top” does not provide services to a higher layer. The seven functional layers are provided by a combination of applications, operating systems, network card device drivers, and network hardware that allow a system to transmit a signal over an Ethernet or fiber optic cable on the network or over Wi-Fi.
What are the layers in the OSI model?
The ISO-OSI layer model consists of a total of seven layers: the Physical Layer, the Data Link Layer, the Network Layer, the Transport Layer, the Session Layer, the Presentation Layer, and the Application Layer. A small tip: The first letters of the individual layer names can be easily remembered by using the phrase “Please Do Not Throw Salami Pizza Away.”
For better understanding, a short overview of the individual layers and their tasks is provided below:
Layer 7 (Application):
Provides functions for applications and data input and output
Layer 6 (Presentation):
System-dependent data is translated into an independent format
Layer 5 (Session):
Manages connections and data exchange
Layer 4 (Transport):
Data packets are assigned to an application
Layer 3 (Network):
Routing or “delivery” of data packets to the next communication node
Layer 2 (Data Link):
Segmentation of packets into “frames” and addition of checksums
Layer 1 (Physical):
Conversion of the received bits (information) into a signal suitable for the medium; finally: physical transmission
Note regarding this: The devices and the transmission medium are not included in the OSI layer model. Nevertheless, in practice it can occur that the devices are specified on the application layer and the transmission medium on the physical layer.
How does the OSI model work in practice?
To ensure that the Open Systems Interconnection model works smoothly in practice, the different transmission and switching technologies and model protocols are assigned to individual layers. What is important here is that it is by no means a “rigid” concept in which there is only one suitable layer for each protocol. It is rather the case that many of the protocols and transmission methods use several layers for their purposes. In practice, it can look like this:
Which protocols work on which layers?
- Layer 7 (Application): Telnet, FTP, HTTP, SMTP, NNTP
- Layer 6 (Presentation): Telnet, FTP, HTTP, SMTP, NNTP
- Layer 5 (Session): Telnet, FTP, HTTP, SMTP, NNTP, NetBIOS, TFTP
- Layer 4 (Transport): TCP, UDP, SPX, NetBEUI
- Layer 3 (Network): IP, IPX, ICMP, T.70, T.90, X.25, NetBEUI
- Layer 2 (Data-Link): LLC/MAC, X.75, V.120, ARP, HDLC, PPP
- Layer 1 (Physical): Ethernet, Token Ring, FDDI, V.110, X.25, Frame Relay, V.90, V.34, V.24
All protocols to which programs and applications have direct access are defined in application layers 5, 6, and 7. In the Windows network world, SMB establishes the connection to the transport layer with NetBIOS. If a Unix or Linux operating system comes into contact with Windows, for example, the Samba service is used to enable proper data transfer. The connection between the application and transport layer is established via TCP ports. Applications and services receive their data directly and automatically via these ports. The data stream is wrapped into “packets” by the connection-based transport protocol (TCP) or the connectionless transport protocol (UDP). The Internet Protocol (IP) then takes over, metaphorically speaking, addressing the packets.
How do programs communicate within the application layer?
In the Application layer, the URL or Universal Resource Locator (on Windows networks the NetBIOS name of the computer) is used to identify a computer and the services running on it. The hosts application, which lists all URLs and IP addresses, is often used to convert the URL into an IP address. Since there are, of course, a lot of URLs, the DNS (Domain Name System) was introduced, which is hierarchically structured for the sake of simplicity. DNS servers are able to request initially unknown DNS names from the parent DNS server. On a Windows network, the lmhosts application or WINS (server) is used to convert NetBIOS names into readable IP addresses. If the latter is then “resolved” in the Transport layer, ARP (Address Resolution Protocol) is used to resolve the IP address into the MAC address (Media Access Control) of the network card (Physical Layer). Note: The MAC address is the only definitive address that can be used to reliably identify a computer on the network, and it is permanently set on a network card.
Which services work across layers?
Cross layer functions, i.e. services that can affect more than one layer, also include the following:
- Security services according to the ITU-T X.800 recommendation
- Management functions that enable two or more units to configure, instantiate, monitor, and terminate communication
- Multiprotocol Label Switching (MPLS) operates on an OSI model layer, which is between Layer 2 (Data Link Layer) and Layer 3 (Network Layer). MPLS can be used to transmit a variety of traffic, including Ethernet frames and IP packets
- ARP translates IPv4 addresses (Network Layer in the OSI model) into Ethernet MAC addresses (Data Link Layer)
- Domain Name Service: an application layer service used to look up the IP address of a domain name
What are the advantages of the OSI model?
The advantages of the ISO-OSI layer model are as varied as its structure: On the one hand, it is an excellent standard model for computer networking—the respective operating systems and other manufacturer-specific differences between devices that want to be networked are of no relevance. Both connectionless and connection-based services are supported which is a big advantage because users can easily use connectionless services, for example when they need faster data transfer over the internet. The connection-based model is still available and can be used when reliability is critical and you don’t want to be dependent on Wi-Fi. In addition, OSI has extraordinary flexibility to adapt to many different protocols. Taken as a whole, this makes it more adaptable and secure than bundling all services in just one of the seven layers.
The disadvantages of the OSI model
In order for the OSI model to work as described, despite the greatest possible freedom with regard to the devices used, “all stakeholders” must nevertheless adhere to certain rules of the game. These rules are defined in the model protocols, which apply to the individual layers or are used across layers. And although the system cannot manage without these protocols, they are by no means clearly defined. Furthermore, some layers are not as useful as others: For example, the layer responsible for session management and the layer (Presentation) dealing with user interaction are not as useful as other model layers. Another drawback is that the respective layers cannot work concurrently or separately—they are reciprocal, meaning that each layer must “wait” to receive data from the previous layer before it can do its job. And this is a common theme running through the entire model. Without this interaction it would not work properly. In addition, some services are duplicated on different layers (for example, the Transport Layer and the Data Link Layer (backup), but without being redundant—and each has its own error control mechanism.
What other layers are there?
Even if only layers 1 to 7 are officially defined in the model, layer 0 is generally used for the cabling and hardware (the “Start Layer” so to speak). Layer 8 represents the user and his requirements. Thus, a Layer 8 error is, for example, an incorrect use by the user and a defective network cable or similar is a Layer 0 error.
The OSI model vs. the TCP/IP model
Like the ISO-OSI layer model, the TCP/IP model is a solution that describes the functions of a telecommunications or network system and, in simple terms, represents a set of protocols for connecting and communicating between network devices. Both are logical models, provide a framework for the creation and implementation of network standards and devices, and serve to divide complex functions into simpler components. One of the biggest similarities between the two systems is their layered structure. However, OSI uses seven layers, while TCP/IP uses only four.
In addition, the top layer represents the Application Layer in each model; it fulfills the same tasks, which, however, differ depending on information content.
But what exactly are the fundamental differences between the two models?
While OSI uses three layers (Application, Presentation, and Session) to define the functionality of the upper layers, TCP/IP uses only one layer (Application) for this purpose
OSI uses two separate layers (Physical and Data Link) to define the functionality of the lower layers, while TCP/IP uses only the Data Link Layer
OSI uses the Network Layer to define the routing standards and protocols, while TCP/IP uses the “Internet Layer”
Since almost all networks today run on the basis of TCP/IP, this model is also taken into account in the OSI model. However, many TCP/IP protocols and transmission methods use more than one layer and thus extend over several OSI layers, in many cases making an exact assignment practically impossible.
The OSI model: what you need to know
The OSI model, often referred to as the ISO-OSI layer model, is a reference model that can be used to describe and define communication between systems. It has seven individual layers, each with clearly delineated tasks. Learn how Myra Security’s solutions protect the most relevant layers (3, 4, and 7) from malicious traffic and attacks.
If you are interested in futher informations, we are willing to send you our product sheet for free
How Myra DDoS Protection can reliably secure your website or web application against all DDoS attack vectors:
- How is the protection activated in case of attack?
- What are the advantages of Myra protection solution?
- What are the features of Myra DDoS Protection for web applications?