Session
The session layer of the seven layer model is where the first basis of communication happens. Session is considered a form of management utility and makes sure that the differing ports and devices can talk and listen to each other. The session layer technology uses a programming language called API, Application Programme Interface which allows programmers to develop network applications.
Session works with half duplex or duplex only as the information to be sent needs to be bidirectional, something that is not possible with simplex communication. For all sessions to be productive there must be something to control them, this is the session layer. The session layer utilises many protocols, namely that of the domain name services, this is a procedure of conversion. It converts the actual webpage in to an I.P address, and secondly the remote procedure calls which is a protocol used by a program to request a service from a program located in another computer in a network. RPC uses what is known as the client/server model.
Session is only considered the instigator of communication as it plays no part in the transporting of data that comes in the next layer called transportation. This shall be described below.
Transportation
The transportation layer is considered to be part of the transportation instruction set of the seven layer model. This can be argued about though. The layer has many similarities with the application set. The transportation level has a vast job of linking the bottom and top half of the seven layer model and has a number of duties when considering that its main task overall is to link and enable communications between differing software applications on different computers. Computers are multitasking and can run many programmes at a time. We can send and receive lots of information, but how does the computer decipher this from differing software? This is the role of the transportation layer.
When sending information the transportation has to split the information up in to smaller chunks, it does this to make the data throughput more efficient and it aids greatly in error detection. This process is reversed when taking the data, keeping track of what comes in, and then combining it to make the whole data packet again. The transport process also includes many methods and functions to increase maximum potential in error sending and receiving. It uses algorithms to ensure the sending of data occurs at a fast and efficient rate. As the layer of the model uses error detection it posses something which is known as connection orientated processing. This method controls the loss of information when transmitting and sends those that failed again. The main protocol used in this stage is the transmission control protocol, TCP. This protocol is normally combined with the IP address and results in the following, it allows information to be sent in packets. The IP handles the actual delivery of the data where as the TCP keeps track of the smaller packets so that information does not get lost on the way. This makes up a huge part of the transportation process however differing protocols to TCP can be used. TCP/IP is one of the most common procedures in computing communications and plays a big part in transportation. This goes to show the relationship between the network layer and the transport layer. The network layer is the IP and the TCP is part of the transport layer.
There are many specific ways in which the transportation layer of the model works and these shall be listed below;
- Segmentation, packaging and reassembling – This process is very similar to what occurs in the memory management sub system of an operating system. Large amounts of data are segmented into smaller pieces on the source machine and are then sent, this is then reassembled on the destination machine.
- Flow control – This is implemented to ensure that time differences between sender and receiver are dealt with and corrected appropriately.
Network
The network layer is where a router will reside and is the area in which differing protocols reside. It is possible to think of the network layer of doing the following task; if two differing networks are involved in communicating then certain pathways need to be followed, ultimately this is the routers job but it co-insides with the network layer. The most common form of protocol used in this stage is the IP described in the previous stage. For the computers to be able to communicate there needs to be a form or a method that is conformed to, this conformation process is called logical addressing. Logical addressing uses I.P addresses assigned to the various computers to locate the machines and send the data to it. From here it is important that the router does its job, moving across a series of networks to send the information to its required address.
On this level we come across basic forms of error detection, this is an important part of the network layer. The routers main responsibility is to route the information so it gets to its target the same as when it had left. To do so error detection is implemented. To do this special protocols are implemented. This now allows the information to be passed down to the data link layer. The data link layer has a special property, and sometimes dictates a length size of information. This then means that if a message is to long it must be spilt up into smaller packets by the networking layer and sent to the data link. This is then reassembled at the other end by the receiving machine.
Data Link Layer
The data link layer is the place in which Local area network technologies reside. By this I mean that technologies such as wireless fidelity, Ethernet, token ring, topologies such as bus, star and ring all occur. This then gives the data link term the tag of a simple network. The data link from here can thus be divided in to two more categories’ which are, the logical link control (LLC) and the media access control (MAC). The purpose of doing this is to intentionally split the two, therefore making differing technologies interpretabilities far easier. The two differing sections thus have their own specific functions and controls and these shall be briefly explained below.
Logical Link Control (LLC) – This is in charge of the control of logical links between local devices on a network
Media Access Control MAC – This is the procedure used by devices across a network to ensure that no conflicts occur. Conflicts generally occur due to shared connection devices, such as a single network and so on. Ethernet is a prime example of this and uses a technology called CSMA/CD.
When sending across the network the data link layer is ultimately the last to brand data packets with addresses of where the information is to go. These addresses normally are MAC addresses. As this is the last layer before the information is actually sent it is important that the labelling of addresses is done correctly as this is where it will finally go. This information is also assigned an algorithm that can detect errors and will ultimately be what is instrumental in ensuring efficient data sending. This is called a cyclic redundancy check (CRC). From here the data link layer can also place packets into frames and this can be beneficial for the following reasons;
- Higher data throughput
- Better error detection
- Ease of unpacking at recipients end
It is this stage that bridges and switches are commonly found. However these are not the only technologies we have when looking at layer two of the seven layer model. Network interfacing cards, also known as NIC are closely related to layer 2 of the model. This is due to the network layer being so close to the actual physical layer of the system.
Physical Layer
This is the final stage in the long process of transmitting data or media. This is considered a raw data form and contains only bits. This is then transmitted using electrical signals. The most uncommon thing that this layer has now to any other of the layers is that this one does not directly connect to the TCP/IP protocol. It is also different because it is the only one that actually sends data physically. It is important to branch away from the idea however that the physical layer is just about hardware devices such as cables routers hubs and switches. This is not the case.
The data link layer is the final labelling control unit but the physical layer is the one that has to interpret this information. This means it is vital in defining and understanding the differing hardware components and specifications in order for the data to be transmitted accordingly. To send the information it must be done in a signal form, this is done by logic gates, the logic gates transform the bits and bytes into signals and are the ones responsible for the sending of the data, this is called encoding and signalling. When this process has been undertaken then data transmission and reception occurs. This happens for both sending and receiving and also occurs in wireless networking. The devices such as the routers, hubs and modems are the ones responsible for possessing the data. They don’t intend on knowing any of the message being sent they are just interested in the raw data, although instead of viewing this as a 0 or a 1 they see it as a light pulse or electric charge.
The two layers, physical and data link are so closely related that it is impossible to define hardware that belong to one or the other. What is tending to happen nowadays is that the physical layer is being divided in to sub categories so that different network Medias can be supported by the same technology, this allows them to increase in performance and why we have seen the following, Ethernet, Fast Ethernet and 10-Gigabit Ethernet.
Conclusion
In concluding it is important to remember that all stages of the seven layer model are equally important even if some functions overlap. They are vital in the way in which we communicate and are always developing to suit new technologies and hardware. Another thing to consider is the actual definition of the seven layers. This is not true as really the system is an eight layer concept where the eighth layer is actually the network engineer or user. In some cases the eighth layer is used to define the way in which the system is protected so it represents the security features.