Advance in computing technology especially the wide area networks (WAN) created a computer networking revolution as companies realized that they could save much money and increase productivity by using network technology. Hence, as new hardware and software with different implementations were introduced, they were added to the expanding networks. As such, networks using different specifications and protocols failed to communicate with each other effectively and thus became incompatible. To address this problem and create an interoperable network, the International Organization for Standardization (ISO) researched network schemes and came up with the Open Systems Interconnection (OSI) reference Model that was released in 1984. The OSI model was officially known as ISO Standard 7489 (Goldman & Rawles, 2004).

The ISO recognized that there was need to create a network model that would help vendors create interoperable network implementations. The OSI Model thus became the primary architectural model for inter-computer communications. Even though other network architectural models have been developed, most network vendors relate their network products to the OSI model when they want to educate users about their products’ compatibility and interoperatability between the various network technologies available worldwide. Therefore, the OSI model has been come the best available tool for learning network technology (Amato, 2000).

The OSI model divides the problem of moving information between computers over a network medium into seven smaller manageable networking functions called layering (layers). The layers 1 through 3 of the OSI model are often referred to as media layers because they control the physical delivery of messages over the network. While the layers 4 through 7 are referred to as host layers and they provide the most accurate delivery of data between computers. Dividing the network into layers has several advantages: that is, it defines the standard interfaces for plug and play compatibility and multi-vendor integration; enables engineers to specialize design that can interoperate; prevents changes in one area from affecting other areas, so each can evolve more quickly; and it divides the complexity of internetworking into discrete and more easily learned operation subsets, etc.

The OSI model architecture defines the communications process as set of the seven layers with specific functions isolated to and associated with each layer discussed as follows:  The Layer 7, which is the application layer, is the OSI model layer closest to the user that provides network services to user applications. The application layer identifies and establishes the availability of intended communication partners, synchronizes cooperating applications, and establishes agreement procedures for error recovery and control of data integrity. It also determines whether sufficient resources for the intended communication exist. The application layer differs from other OSI layers in that it does not provide services to them but rather application processes outside the scope of the OSI model.

Layer 6 is the presentation layer, which ensures that information sent by the application layer of one system is readable by the application of another system. If necessary, the presentation layer translates between multiple data representation formats by using a common data representation format. For example, the presentation layer can perform ASCII-tot-non-ASCII character conversions, encryption and decryption of secure documents, and the compression of data into smaller packets (White, 2004).

The session layer establishes, (also known as Layer 5) manages, and terminates sessions between applications. It synchronizes dialogs between two or more presentation entities and manages their data exchange. That is, the session layer provides services to the presentation layer. The synchronization points established often act as backup points in case of errors or failures. For example, when transmitting large documents, the session layer might insert synchronization points on each page. If an error occurs during transmission, both the sender and receiver can back up to the last synchronization point and that becomes the start point when retransmission resumes. Some network applications do not include a specific session layer, hence, if network application uses synchronization points or token sessions, these points or token sessions are inserted by the application layer or possibly by the transport layer (White, 2004).

Layer 4 is the transport layer, which segments and reassembles data into a data stream. It determines how reliable data transfer is over an inter-network and provides mechanisms for the establishment, maintenance, and orderly termination of virtual circuits, transport fault detection and recovery, and information flow control to prevent one system form overwhelming another with data (Amato, 2000). The transport layer protocols that facilitate the flow of data are transmission control protocol (TCP), which is a connection-oriented protocol that provides reliable end-to-tend data transfer capability; and the user datagram protocol (UDP), which is connectionless protocol that depends at times upon higher layers in the protocol suite for reliability (Held, 2001). The transport layer operations are very similar to data link layer operations. There main difference is that the transport layer performs only at end points while the data link layer performs it operations at every stop (node) along the path (White, 2004).

Layer 3, also known as the network layer, is the most complex layer that provides connectivity and path selection between two end systems that may be located on geographically diverse networks. Thus routing data through the network onto the correct paths is an important feature of this layer. Several protocols have been defined for the layer 3 to include the CCITT X.25 packet switching protocol, which governs the flow of information through a packet network, and the CCITT X.75 gateway protocol, which governs the flow of information between packet networks (Held, 2001).

The data link layer (Layer 3) provides reliable data transmission across a physical link. This layer deals with the physical addressing, network topology, line discipline, error notification, ordered delivery of frames, and flow control. The data link layer also provides services to the network layer. Since the development of OSI layers was originally targeted toward WAN, its applicability to local area network (LAN) required a degree of modification. Under the IEEE802 standards, the data link was originally divided into two sublayers: logical link control (LLC) and media access control (MAC). The LLC layer is responsible for generating and interpreting command that control the flow of data and perform recovery operation when an error occurs. While the MAC provides access to the LAN, thus enabling stations to transmit information. Data link control protocols such as binary synchronous communications (BSC) and high-level data link control (HDLC) reside inside this layer  (Held, 2001).

Layer 1 is the physical layer that defines the electrical, mechanical, procedural, and functional specifications for activating, maintaining, and deactivating the physical link between two systems (Goldman & Rawles, 2004). Such characteristics as voltage levels, timing of voltage changes, physical data rates, maximum transmission distances, physical connectors, and other attributes defined by the physical layer specifications. The physical layer also provides services to the data link. Typically, the physical layer corresponds to established interface standards such as RS-232 (Amato, 2000 & Held, 2001).

Taking a critical look at the OSI model, the layers do not specify precise protocol or exact services. For instance, the network layer is concerned with finding the best path for data transmission from on point to another, but it relies on the transport layer to make sure that the data received at the final destination is the same as the original data. Likewise, the application layer or transport layer can sometimes play the role of the session layer; therefore, rendering the session layer insignificant. Thus, it should be noted that the OSI model is not a network implementation tool but it rather specifies the purpose and functions of each network layer.


Reference:

Goldman, E.J. & Rawles, T.P. (2004). Applied Data Communications: A Business-Oriented Approach. Wiley

Held, G. (2001). Understanding Data Communications: Wiley

Amato, V. (2000). Cisco Networking Academy Program: Companion Guide. Cisco Press.

White, C. (2004). Data Communications & Computer Networks: A Business User’s
Approach. Thompson Course Tech