Introduction
This report will cover the differences and similarities of the architecture and PDU of both the OSI model and the TCP/IP model. These framework designs allow data to be transferred by splitting protocols into layers. This affects the software and hardware and the hardware. This report will also discuss the history of the layers OSI and TCP/IP models and the company and employees who created these models. This report will also include the advantages and disadvantages as well as the most efficient model for a smaller or larger network taking into consideration the protocols that are associated with the layers and how they are distributed.
The OSI Model
Brief History
The OSI (Open System Interconnection) model was developed in the late 70’s by the organisation “ISO” together with “The International Telegraph and Telephone Consultative Committee” (CCITT). Both the organisations created documents that defined similar network models. In 1983, the two documents were merged to form “The Basic Reference Model for Open Systems Interconnection”. It was published to the public by both companies in 1984. Various aspects of OSI model have been changed from the experiences between the various companies. The new design was documented in ISO 7498. In this model, a networking protocol system was sectioned into layers. Within the layers, each of them had one or more objectives to implement its use. Each objective is able to interact directly with the layer immediately beneath it and provides the facilities to be able to be used by the layer above it.
The Architecture (Layers)
The OSI model contains 7 layers of architecture for the distribution of data transfer over a network.
Physical – This layer lowest layer which provides both mechanical and electrical signals. Electrical signals can be transferred through mediums such as optical fibre, laser or electromagnetic waves (wireless networks) or sound. There are some devices and network components that involve themselves within the physical layer, these include hubs, cabling, repeaters, sockets and antennas. This layer is responsible for the transmission and receipt of unconstructed raw data in a physical method. The physical layer is also responsible for bit rate control as well as the transmission modes including simplex, half duplex, and full duplex. Simplex is when transmissions can only be sent one way. Half duplex is when it is possible to send both directions, but, it can only send or receive one way at a time. Full duplex is the best and most common which allows for transmissions both ways simultaneously. The physical layer is also layer responsible for defining the network topologies, (bus, mesh or ring). This is based on the hardware within the network, for example, network adapters, repeaters, network hubs, modems, and fibre media converters. Fast Ethernet, RS232 and ATM are protocols associated with the physical layer components.
Data Link – This layer is the second lowest layer and makes sure that a reliable transfer of data is possible between the source or sender and the user or receiver is obtainable. Flow control makes it possible to control the rate/speed to avoid collisions. It further allows for node-to-node transfer which is a link directly between two connected nodes. This can detect and correct any errors that occur on the physical layer (lowest layer). This layer contains the protocol for flow control to be used between two physically connected devices. This layer also contains two sublayers known as, MAC (Media Access Control) which allocates an address by controlling the method that a device is able to access the network and transmit data as well as LLC (Logical Link Control) which allows recognition and encapsulation of the network layer protocols. The LLC also controls sublayers to allow error checking and frame synchronisation to take place.
Network – This layer is the third lowest layer and provides the method for variable length data sequences known as ‘datagrams’. Datagrams are then transferred from one node to another in a separate network. A network is a medium which allows many nodes to allow communication between one another to complete a connection. Every node has an address and which allows the nodes connected to be able to transfer messages from one another when they have an established connection. This merely provides the content of a message and the address of the correct destination node. By letting the network find the directions to deliver the message to the destination node, possibly routing it through intermediate nodes. If the message required to be sent is too large to be transmitted from one node to another via the data link layer between two nodes, the network may implement the message delivery by dividing the message into several smaller fragments at one node, sending the fragments independently to the second node, thus reassembling the fragments at the second node to recreate the full message. The system may automatically report an error in some cases.
Transport – This layer is the fourth lowest layer and provides the means necessary of being able to transfer variable length data sequences from the source to the destination host required by the user, via one or more network channels while maintaining the quality of service to fulfil the functions. This layer controls the reliability of a given address through flow control, segmentation, de-segmentation and error control. Some protocols are orientated by the state- and connection. This allows the transport layer to keep track of the segments and retransmit those that fail throughout their course to the destination host. The transport layer also provides the acknowledgement of the successful data transmission and sends the next data if no errors occur. The transport layer is tasked to create packets that made out of the message received from the application layer.
Session – This layer provides the connections between computers. This layer is tasked to establish, manage and terminate the connections between the local and remote applications. It is able to provide capability to allow for full-duplex, half-duplex, or simplex operation connections, and establish checkpoints, flow control, the ability to terminate, and restart procedures when required. The OSI model made this layer responsible for closure of sessions, which is one of the properties of the Transmission Control Protocol, and for session checkpoints and recovery. The session layer is commonly required to be implemented explicitly in application environments that use remote access procedures.
Presentation – This layer creates the framework between application-layer units, in which the application-layer allows the use of different syntax within the presentation layer. This provides a mapping between them. If a mapping is available to be used, the presentation layers data units are encapsulated into session protocol data units which is then passed down the protocol stack. This provides independence from data representation by translating between application and the network formats used. This layer also is tasked with transforming the data into a form that is readable and accepted by the application layer. This layer is tasked with formatting the data to be sent across a single network. This process is in some cases called the syntax layer. The presentation layer also includes compression functions. This layer is also tasked with negotiating the transfer of syntax.
Application – This layer of the OSI model is the closest to the end user, which means both the OSI application layer and the user interact directly with the software application protocols. This layer is tasked with interacting with software applications that implement a component for communication. The Application layer functions include identifying partners for communication, allocating resource availability, and the synchronisation of communication over a network. When this layer is tasked with identifying communication partners, the application layer determines the identity and availability of communication partners for an application to be able to transmit data signals. The most important feature within the application layer is the ability to create a distinct difference between the application-entity and the application itself. For example, a reservation website might have two application-entities: one using the HTTP protocol to communicate with its users, and another protocol for accessing remote database to record reservations. Neither of these protocols has anything to do with reservations. The application layer has is not able to check the availability of resources in the network.
The TCP/IP Model
Brief History
The Internet Protocol Suite, which is most commonly known as TCP/IP (Transmission Control Protocol/Internet Protocol), was created by the Defence Advanced Research Projects Agency (DARPA) within the late 1960’s. In 1972, Robert E. Khan joined DARPA where he worked on satellite and ground-based packet networks and the ability for communication between both the satellite and the ground packet networks. In 1973, Vincent Cerf joined Kahn in creating this idea and was later completed later within that year and it was published in 1974, the year after to be known as TCP (Transmission Control Program). Both Kahn and Vincent are now known by the world as the developers of the internet. DARPA then contracted with BBN Technologies, Stanford University, and the University College London to develop operational versions of the protocol on different hardware platforms. The four versions that have been developed are known as TCP v1, TCP v2, TCP v3 and IP v3, and TCP/IP v4. The Protocol TCP/IP v4 is still in use today at this current point of time. As the protocol grew, scalability became a problem and so the Transmission Control Program was then sectioned/divided into two separate protocols. These protocols are formally known as the Transmission Control Protocol and the Internet Protocol (TCP/IP). The TCP/IP protocol has also been a standard for military computer networks within the US department of defence since 1982.
The Architecture (Layers)
The TCP/IP model contains 4 layers of architecture for the distribution of data transfer over a network.
Application – This layer includes 4 different protocols which are known as, HTTP (Hypertext Transfer Protocol)/HTTPS (Hypertext Transfer Protocol Secure), FTP (File Transfer Protocol), SMTP (Simple Mail Transfer Protocol), DHCP (Dynamic Host Configuration Protocol). The protocols discussed above are formally used within most applications to be able to provide user services or for exchanging application data over the network connections which are established by the lower level protocols. This layer within the TCP/IP model is often compared as equivalent to a combination of the session layer, presentation layer, and the application layer within the OSI model.
HTTP – the purpose of the HTTP protocol is to have a standard method for communication of transmitting data across the internet between network connected devices and applications. This includes devices and applications such as servers, web browsers and other network devices, utilities or applications. By default the HTTP protocol uses port 80. A web server uses the port number to allow access for the service that is required by the client. The port numbers can be found from the list of well known port numbers from port 0 – 1023 and for registered port numbers from port 1024 – 49151. The OSI and TCP/IP suite both have link in common within the application layer, whereas, TCP/IP combines the first three layers that are within the OSI model to allow the creation of the application layer.
HTTPS – the purpose of the HTTPS protocol is to have a standard method for communication of transmitting data across the internet between network connected devices and applications which is secure. HTTPS uses one of the two protocols to allow for the end to end connections to be encrypted. These protocols are known as the SSL (Secure Socket Layer) and TLS (Transport Layer Security). This includes devices and applications such as servers, web browsers and other network devices, utilities or applications. By default the HTTPS protocol uses port 443. A web server uses the port number to allow access for the service that is required by the client. The port numbers can be found from the list of well known port numbers from port 0 – 1023 and for registered port numbers from port 1024 – 49151. The OSI and TCP/IP suite both have link in common within the application layer, whereas, TCP/IP combines the first three layers that are within the OSI model to allow the creation of the application layer.
FTP – the purpose of the FTP protocol is to have a standard method for the transfer of data across the internet between network connected devices and applications. This includes devices and applications such as servers, web browsers and other network devices, utilities or applications. By default the FTP protocol uses port 20 for the client end and port 21 for the server end. A web server uses the port number to allow access for the service that is required by the client. The port numbers can be found from the list of well known port numbers from port 0 – 1023 and for registered port numbers from port 1024 – 49151. The OSI and TCP/IP suite both have link in common within the application layer, whereas, TCP/IP combines the first three layers that are within the OSI model to allow the creation of the application layer.
SMTP – the purpose of the SMTP protocol is to have a standard method for the sending and transferring of mail from the source to the destination. This can be process can be utilised by the end user to send mail as well as for use of transfer between mail servers. This process is utilised by devices and applications such as servers, web browsers and other network devices, utilities or applications. By default the SMTP protocol uses port 25. A web server uses the port number to allow access for the service that is required by the client. The port numbers can be found from the list of well known port numbers from port 0 – 1023 and for registered port numbers from port 1024 – 49151. The OSI and TCP/IP suite both have link in common within the application layer, whereas, TCP/IP combines the first three layers that are within the OSI model to allow the creation of the application layer.
DNS – the purpose of the DNS protocol is to have a standard method for translating web addresses (domain names) to their relative ip addresses and subnets. This helps the end user by them not having to memorise the ip addresses and subnets to open the required web page they individual wishes to open by only memorising the domain name. This process is utilised by devices and applications such as servers, web browsers and other network devices, utilities or applications. By default the DNS protocol uses port 53. A web server uses the port number to allow access for the service that is required by the client. The port numbers can be found from the list of well known port numbers from port 0 – 1023 and for registered port numbers from port 1024 – 49151. The OSI and TCP/IP suite both have link in common within the application layer, whereas, TCP/IP combines the first three layers that are within the OSI model to allow the creation of the application layer.
Transport – This layer establishes connections for basic data channels that applications use to complete task-specific data exchanges. This layer also establishes process-to-process connectivity, this means that it is able provide end-to-end services that are independent of the structure of user data and the logical arrangement of exchanging information for the specific purpose required. This protocols responsibility includes end-to-end message transfer independent of the underlying network conditional agreements, along with error control, segmentation, flow control, congestion control, and application addressing (port numbers). The TCP is a orientated connection protocol that is able to address numerous reliability issues whilst providing a reliable byte stream to makes sure that the data arrives in-order, with minimal errors in the process, duplicate data is discarded in the process, as well as lost or discarded packets being resent to make sure the complete data is received, this includes traffic congestion control.
Network/Internet – This layer is responsible for the sending of packets across multiple networks which requires inter-networking for sending data from the source network to the destination network. This process is called routing. The Internet Protocol performs two basic functions: Host addressing and identification and packet routing which is the basic task of sending datagrams from source to destination by forwarding them to the next closest network router to the destination required. The internet layer protocol provides an unreliable datagram transmission capability between hosts located on potentially different IP networks by forwarding the transport layer datagrams to an appropriate next-hop router for further relaying to its destination. By using this functionality, this layer makes it possible for inter-networking to be processed, the interworking of different IP networks, and establishes the Internet. The Internet Protocol is the main component that forms the internet layer, and it is able to define the two addressing systems to identify the network hosts and to locate them on the network. The original address system of the ARPANET and its successor, the Internet, is Internet Protocol version 4 (IPv4). It uses a 32-bit IP address and is capable of identifying approximately four billion hosts. This limitation was removed in 1998 by the standardisation of IPv6 which uses 128-bit addresses. The implementations of IPv6 emerged originally in 2006 and are still being implemented to this day.
Network Interface/Link – This layer is the networking scope of the local network connection that a host is attached to. This management is called the link in TCP/IP literature. It is the lowest component layer of the Internet protocols. TCP/IP is designed to be hardware independent and as a result, TCP/IP can be implemented on top of the hardware technology of any network. The link layer is used to allow the transfer of packets between the Internet layer interfaces of two hosts on the same link. The process of transmitting and receiving packets on a given link is able to be controlled both in the software device driver/s for the network card, as well as within firmware or the chipsets. These allow access of data link functions to be performed such as adding the specific required packet headers for preparation of the packets to be transmitted from the source, they then transmit the frame over a physical medium. The TCP/IP model includes specified methods for translating the network addressing methods used in the Internet Protocol to link the layer addresses, such as MAC addresses.
Comparisons – (Similarities and Differences)
Both of the models (TCP/IP and OSI models) assume that packets are able to be switch and thereby allow each of the individual packets to create their own paths to reach their required destinations. The Data is added to the header within the packet to allow it to direct the packet towards its destination where the data contained after the packet contents are extracted (de-encapsulated). Packet switching is used for the primary basis for data communications in the world wide web and in computer networks worldwide.
The transport service in both models is able to provide a solid and reliable end-to-end byte stream. This means that the destination device receives the file or data in exactly the same amount of storage size and correct order as they were sent. The protocol for this is TCP.
The OSI and TCP/IP model both include a common application layer, whereas, TCP/IP combines the first three layers that are within the OSI model to allow the creation of the application layer. This means that every protocol within the application, presentation and session layers that are based within the OSI model are also based within the application layer of the TCP/IP model. As well as sharing a similar application layer, they also commonly share the transport and internet/network layers which include all the same protocols that are required by both models.
Also, their last layers are similar, OSI has a data link and physical layer which include the same protocols as the same link layer on the TCP/IP which is a lot simpler. So just like the application layer where the first three layers of the OSI model were part of the application in the TCP/IP model, the data link and physical layer protocols are part of the link layer in the TCP/IP model.
Lastly, both models are based on layer protocol.
The TCP/IP model is also more modern and simplistic due to it only consisting of 4 architectural layers as to the OSI model where it consists of 7 architectural layers. As well as TCP/IP is a more credible model. This is mainly due to the fact because TCP/IP protocols are the standards around which the internet was developed therefore it mainly gains creditability whereas networks are not usually built around the OSI model as it is mainly used as a guidance tool. The US Department of defence has also stated that they use the TCP/IP model as the standard for all military networks. This shows that this model has a much higher reliability and stability when it is compared to OSI model.
Reference / Bibliography
https://en.wikipedia.org/wiki/Internet_protocol_suite
http://www.omnisecu.com/tcpip/tcpip-model.php
https://en.wikipedia.org/wiki/OSI_model
http://www.pearsonitcertification.com/articles/article.aspx?p=1868080