The Open Systems Interconnection model (OSI model) is a seven-layer abstraction model created in 1984 by the International Organization for Standardization (ISO) to help vendors create compatible and interoperable networks. The OSI model is a layered model that assists in describing the functions of a communications system and how protocols relate to one another.
A layered approach was used because it reduces complexity and facilitates modular engineering. Each layer serves the layer above it while at the same time depending on the one below. As we go through a brief introduction to the OSI model we will start at the bottom and work our way up.
The physical layer is the first layer of the OSI model. The physical layer defines the electrical, mechanical, and procedural specifications for transmitting and receiving raw data bits over the transmission medium. At this layer, devices strictly deal with the physical signal only and do not concern themselves with any information that makes up the signal itself. Cables, connectors, terminations, frequencies, modulation techniques, etc. are specified at this layer.
Data Link Layer
The data link layer is a protocol layer that is responsible for data transfers between adjacent nodes in the network. Media arbitration also takes place at this layer as multiple devices contend for the shared network medium in order to transmit their data. The data link layer may also be able to detect or even correct errors that result from problems with the physical layer. Hardware addressing is a service provided at this layer to uniquely identify nodes on the local network segment. Ethernet is an example of a layer-2 protocol.
The network layer is responsible for message forwarding and host addressing. The network layer transfers packets of information between hosts on different networks, and will choose the best path for the packets to travel as they go from source to destination. This layer is considered connectionless because the destination host doesn’t send any sort of acknowledgement of receipt of the packet. Logical addresses are used to assign a unique address to each host on the network. Internet Protocol (IP) is an example of a layer-3 protocol.
The transport layer ensures end-to-end connectivity for hosts. Fault detection and recovery are key services at this layer. Flow control is also used at this layer to ensure that a host is transmitting at an acceptable rate (not too fast) for the receiver. The transport layer also ensures that fragments of a message received by a host can be reassembled in the original order, as this layer cannot guarantee that the data will arrive in the order that it was sent. Transmission Control Protocol (TCP) is an example of a layer-4 protocol.
The session layer establishes, maintains, and tears down sessions between applications. This layer may try to recover a dropped connection in the event of a connection loss. Remote Procedure Call (RPC) is an example of a layer-5 protocol.
The presentation layer arbitrates data transfer syntax for the application layer. This layer also focuses on data format and can provide data conversion services. Data encryption/decryption are also important services that can be offered here.
The application layer is the closest to the user and provides a means for the user to access information and resources on the network by way of an application interface. Both the user and the application layer interact with the software application. DNS, HTTP, and email (SMTP) are examples of layer-7 protocols.
At first glance, it is easy to write off the OSI model as a little too “fluffy” and academic – something only created for the purposes of exams and not having any value in the real world. This would be a mistake. The OSI model is an excellent teaching aid and even a useful reference model for troubleshooting. In fact, one could argue that describing the functions of the various equipment found in a network to be nearly impossible without referencing the OSI model.