OSI model

OSI Model

 

The ISO (International Organization for Standardization) decided to construct a framework of standards in which different vendors would be able to use in order to communicate over a network consisting of diverse equipment and applications.  This framework is now considered the standard for communication of networks.  The OSI is divided into 7 layers, which divides the task into smaller more manageable task groups.  Each task or group of tasks assigned to each layer can also be implemented independently.  This limits complications between layers because the solutions offered by one layer do not adversely affect the other layers.

The 7 layers can be split logically into two subgroups.  Layers 7 thru 4 focus on the end to end communication of data source and destinations.  Layers 3 thru 1 are provide consistent communication between the network devices.  An easier way of looking at the OSI model is dividing the upper layers (7, 6, 5) from the lower layers (4, 3, 2, 1).  The upper layers deal with application issues and are implemented only in software.  The highest layer, the application layer, is the closest to the end user.  The lower layers are responsible for the transportation of the data.  The physical layer and the data link layer are implemented in hardware and software. The lowest layer, the physical layer, is closest to the physical network medium (the wires, for example) and is responsible for placing data on the medium.

The following is a top-down explanation of the OSI Model. It starts with the user's PC and it follows what happens to the user's file as it passes though the different OSI Model layers. The top-down approach was selected specifically (vs. starting at the Physical Layer and working up to the Application Layer) for ease of understanding. It is used here to show how the user's files are transformed (through the layers) into a bit stream for transmission on the network. 

Benefits of the OSI Model

By separating the network communications into logical smaller pieces, the OSI model simplifies how network protocols are designed. The OSI model was designed to ensure different types of equipment (such as network Adapters, hubs, and routers) would all be compatible even if built by different manufacturers. A product from one network equipment vendor that implements OSI Layer 2 functionality, for example, will be much more likely to interoperate with another vendor's OSI Layer 3 product because both vendors are following the same model.

The OSI model also makes network designs more extensible as new protocols and other network services are generally easier to add to a layered architecture than to a monolithic one.

Adapters, hubs, and routers

Physical Layer

At the base of the OSI model is the physical layer. This one is the easiest to understand - it encompasses most of the physical aspects of the network; for example, a repeater (a piece of equipment that amplifies signals) operates at the physical level because it is only concerned with transmitting the electric signal on the wire - it does not try to interfere with, encode/decode, or otherwise logically manipulate the signal. Think of the physical layer as the "electrical" layer of the model - the physical layer is the layer of low-level networking equipment, such as some hubs, cabling, and repeaters . The physical layer is never concerned with protocols or other such higher-layer items.
Examples of hardware in this layer:

• Network adapter
• Repeater
• Network hub
• Modem
• Fiber Media Converter 

 

Data Link Layer

The Data Link Layer transfers data between adjacent nodes in a Wide Area Network, or between nodes on any given Local Area Network. This layer also provides the procedural means to transfer data between network entities and sometimes to detect and correct errors that have occurred within the Physical Layer.
Since the Data Link layer is concerned primarily with local delivery within a LAN, data link frames do not cross the boundaries of a local network segment and instead focus on local delivery, addressing, and media arbitration.
Switches operate in the Data Link layer.

Sublayers

The Data Link Layer is often subdivided into two sublayers, the LLC Sublayer and the MAC Sublayer:

• The LLC (or Logical Link Control) Sublayer multiplexes protocols running atop the Data Link Layer, and provides flow control, acknowledgement, and error control. It also specifies the mechanisms to be used for addressing stations and for controlling the data exchanged between machines.
• The MAC (or Media Access Control) Sublayer determines who is allowed to access the media at any one time (as in CSMA/CD) and provides frame synchronization, which determines where one frame ends and the next begins.

 Protocols

Protocols in the Data Link Layer include:

• Ethernet for LANs
• PPP
• HDLC
• ADCCP for point-to-point connections
• PTPPD Point-to-Point Portal Device

Network Layer 

The Network layer is where the frames of the Data Link layer become packets. It can be described as the puberty of the OSI model. It is where the boys of the Data Link layer become men. The best way to think of the Network layer is as the mailroom clerk of the OSI model. The clerk receives mail and directs it to the appropriate couriers. In similar fashion, the Network layer translates the frames it receives from the Data Link layer into more logical packets which can be routed to other networks (like sending it to a courier). At the Network layer, you can begin to actually communicate across a Network, but the service is called "unreliable" because no connection can be established. Communication over the Network layer is something like throwing a message in a bottle into the sea or hollowing out the wall and writing cryptic references to the Companion Cube at the Aperture Science enrichment center- you cannot verify that the other person ever reads the message. The Network Layer is the layer that uses IP addresses. Most of what we call "routing" occurs at the network layer - that is,network traffic is routed from one network to another at this layer, allowing for inter-network (as opposed to intra-network) communication.

 Transport Layer 

At the Transport layer, the Network layer's packets are sorted and organized into "segments." This is different from the idea of packets in the Network layer in one fundamental way: the segments of data over the Transport layer contain information on the connection and the transmission of data. The Transport layer removes the uncertainty of "throwing the message in the bottle" that we experience over the Network layer by attaching to segments (which are basically continued packets) information about the state of a connection. Thus, the Transport layer uses the LLC sub-layer of the Data Link layer to establish connections between hosts. The protocols that are typically associated with the Transport layer are:

• TCP: Connection-Oriented, reliable - unlike the "message in the bottle," or "hidden Companion Cube reference," it can verify that a segment or packet reaches the location, or note failure
• UDP: Connectionless, unreliable - it is like a "smarter" message in the bottle service; it makes a best-effort delivery but does not establish a connection or verify receipt

Session Layer 

The Session layer is the layer that initiates and terminates the transport layer connection-oriented services . While the Session layer is not widely used by protocols, it is important in that it is responsible for managing the connections that we value and that the Transport layer provides. In other words, the Session layer is like the GLaDOS of the transport layer, just like GLaDOS tells the robots what to do and promises them eternal damnation in case of insubordination, the Session layer also bosses the Transport layer around.

Presentation Layer 

The Presentation layer translates the segments of information from the Transport layer into data that can be used at the Application layer. It is like the C3-PO of the OSI model. It is something of an intermediary between the network node's processing area and the network node's actual networking area - it can interpret the segments or packets it receives and change them into "data formats" that we all know and that the PC can recognize.

 Application Layer 

This is the top of our cake, utilizes the layers below it, and includes the functions that we are most familiar with - the end-user application protocols such as FTP and HTTP, the vital services like DHCP and DNS, and several obscure applications. This is where the sockets are defined. The Application layer does indeed cover a wide variety of protocols and services, but don't let this overwhelm you. In general, when trying to decide if a service or protocol is an application level one, ask: "Does this facilitate networking, or does networking facilitate the service?" If the answer is the latter, you know it is an Application layer service.