OSI reference model

HISTORY:

OSI reference model is advanced teaching of network. As there was a need of common model of protocol layers, this concept promoted the idea and defined about interoperability between network devices and software. Networking with protocol standards such as SNA and Decnet are completely vendor-developed and proprietary which are prior to OSI concept. Industries worked out to sort out such complications, and have introduced a new concept named OSI which is attempting to get everyone to agree to common network standards to provide multi-vendor interoperability. Many devices were unable to communicate with other protocols directly because of lack of common protocols. It was common for large networks to support multiple network suites. Database design and distributed database design were the interest of canepa’s group in the early and middle 70s’. It was clear in the mid 70s’ that standard distributed communications architecture would be required to support database machines, distributed access, and the like. For development of such standardized database systems some of the groups have studied the existing solutions, including IBM’s system network architecture (SNA), so the work on protocols is done for ARPANET, and some concepts of presentation services. In spite of these defaults which lead to the development of a seven layer architecture known internally as the distributed systems architecture (DSA) in 1977.

Meanwhile, in 1977 International Organization for Standardization has got a proposal that a standard architecture was needed to define the communications infrastructure for distributed processing by British Standards Institute. Thus, subcommittee on Open System Interconnection was formed by ISO [Technical Committee 97, Subcommittee 16]. The American National Standard Institute (ANSI) was charged to develop proposals in advance of the first meeting of the subcommittee.

INTRODUCTION:-

Although the topic of OSI can seem boring and academic, actually it is a very useful way to learn more about networking. OSI offers a good starting point for understanding the differences between repeaters, switches, and routers, for example. The model also can help one better understand network protocols.

The main idea in OSI is the process of communication between the end points of a two telecommunication network, this can be divided into layers, which can be added within its layers of its own set of special, related functions. The OSI model (and networking models developed by other organizations) attempts to define rules that cover both the generalities and specifics of networks. OSI model was not a network standard, but it was used as guidelines for development of network standards. It was become so important that almost all major networks standards in use today conform entirely to its seven layers.

The relationship between OSI model and internet protocol stack, as it is known that the new technologies and developments are emerged based on networks. These networks comprise of protocols and the entire protocol stack. Internet protocol stack is based on OSI model. [1]

OSI MODEL:-

The modern day networking process is understood by layering. OSI model consists of seven layers. They are

1. Application Layer (7th Layer)

2. Presentation Layer (6th Layer)

3. Session Layer (5th Layer)

4. Transport Layer (4th Layer)

5. Network Layer (3rd Layer)

6. Data-link Layer (2nd Layer)

7. Physical Layer (1st Layer) [2]

The OSI model defines internetworking in terms of a vertical stack of seven layers. Among, these layer, there are two parts division one is upper layer and the other is lower layer. Upper layer consist of layer 7, 6 and 5 and the lower layer consist of 4, 3, 2 and 1. Among these two layer upper layer deals with the application issues which are generally are implemented only in software. The software that implements network services like encryption and connection management. When we consider deals with the data transport issues and the implement more primitive, hardware functions like routing, addressing and flow control. [3]

Upper layer: –

(Application layer, Presentation layer, Session layer)

OSI designates the application, presentation, and session stages of the stack as the upper layers. Software in these layers performs application-specific functions like data formatting, encryption, and connection management. For example: -HTTP, SSL and NFS

Lower layer: –

(Transport layer, Network layer, Physical layer.)

Lower layers of the OSI model provide more primitive network-specific functions like routing, addressing, and flow control. For example:- TCP, IP, and Ethernet

Application Layer: –

Describes how real work actually gets done. Ex: this layer would implement file system operations. And also it defines interface to user processes for communication and data transfer in network. It provides standardized services such as virtual terminal, file and job transfer and operations.

“ provides service directly to the user application. Because of the potentially wide variety of application, this layer provide a wealth of services. Among these services are establishing privacy mechanism, authenticating the intended communication partners, and determining if adequate resources are present.” [3]

Presentation layer : –

Describes the syntax of data being transferred. Ex: this layer describes how floating point numbers can be exchanged between hosts with different math formats. Mask the difference of data formats between dissimilar systems. Specifies architecture independent data transfer format. It encodes data; like encrypts and decrypts data, compresses and decompresses data.

“ performs data transformation to provide a common interface for user application, including services such as reformatting data compression, and encryption.”[3]

Session layer : –

In general, describes the organization of data sequences larger than the packets handled by lower layers. Ex: this layer describes how request and reply packets are paired in a remote procedure call. It mange user session and dialogues. And also controls establishment and termination of logic between users. Reports upper layer errors.

“ establishes, manages and end user connection and manages the interaction between end systems. Services includes such thing as establishing communication as full or half duplex and grouping data”[3]

Transport Layer:-

In general, describes the quality and nature of the data delivery. Ex: this layer defines if and how retransmissions will be used to ensure data delivery. Manage end-to-end message delivery in network. Provides reliable and sequential packet delivery through error recovery and flow control mechanisms. Provides connectionless oriented packet delivery

“ insulates the three upper layer, 5 through 7, from having deal with the complexities of layer 1 through 3by providing the function necessary to guarantee a reliable network link. Among other function, this layer provides error recovery and flow control between the two end points of the network connections.”[3]

Network layer :-

In general, describes how a series of exchanges over various data links can deliver data between any two nodes in a network. Ex: this layer defines the addressing and routing structure of the Internet. Network layer function determines how data are transferred between network devices. Routes packets according to unique network device addresses. It provides flow and congestion control to prevent network resource depletion.

“ Establishes, maintains and terminates network connection . Among other functions, standard define how data routing and relaying are handled”[3]

Data link layer: –

In general, describes the logical organization of data bits transmitted on a particular medium. Ex: this layer defines the framing, addressing and check summing of Ethernet packets. Defines procedures for operating the communication links and frames packets. And also detects and corrects packets transmit errors.

“ ensures the reliability of the physical link established at layer 1 . standard defines how data frames are recognized and provide necessary flow control and error handling at the frame level.”[3]

Physical layer : –

In general, describes the physical properties of the various communications media, as well as the electrical properties and interpretation of the exchanged signals. Ex: this layer defines the size of Ethernet coaxial cable, the type of BNC connector used, and the termination method. Physical later defines physical means of sending data network devices and defines optical, electrical and mechanical characteristics and interfaces between network medium and devices.

“ controls transmission of the raw bit stream over transmission medium. Standard for this layer define such parameter as the amount of signal voltage swings, the duration of voltage and so on.”[3]

INFLUENCE:

The protocol stack or Internet protocol stack also show influence of the OSI model most of the terminology is same, and most of the books present an Internet protocol stack that uses OSI terminology and that includes physical and data link layer. But in 1981 article by john postal, Carl Sunshine and Danny Cohen there is a diagram showing the Internet protocol functioning on top of one of the several Network protocol and it describe such network protocol as being the packet transmission protocol in each individual network. In a article 1983 by Vinton Cerf- often referred to as “ one of the founding father of the internet” – and Edward Cain that the layer that IP function is called “Internetwork layer” and the layer under that is “Network layer”. We can universally that the layer at which IP function is called the Network Layer and the important layer was called “Network layer is usually omitted.” [4]

TCP/IP: –

TCP/IP was originally developed as a research experiment. It has become the backbone of thr information superhighway. It provides users with the back bone services used to carry popular services such as the World Wide Wed (www), E-Mail and other. In late 70’s as a US Government Advanced Research Project Agency (ARPA) research development project, the “ARPANET”. This has grown to provide connection to system world wide, helping bring the information age into reality.

TCP/IP is not really a protocol, but a set of protocols – a protocol stack, as it is most commonly called. Its name, for example, already refers to two different protocols, TCP (Transmission Control Protocol) and IP (Internet Protocol). There are several other protocols related to TCP/IP like FTP, HTTP, SMTP and UDP – just to name a few.

The TCP/IP suite can be understood as a set of discrete component with a layered architecture. The transport layer and below of the TCP/IP suite can be viewed in the figure. The link covers the internetwork and transport protocols. The delivery of packets between the adjacent nodes in an internetwork id provided bye the link protocols. This link delivery only implies that delivered information is intact. Internetwork layer is responsible for addressing and routing of packets between source and destination nodes within the network. At the final the transport layer provides full service to transfer information between them.[5]

Application layer: –

Application layer is the communication between programs and transport protocol. Several different protocols works on the Application layer. The familiar protocols HTTP, SMTP, FTP, SNMP, DNS and Telnet. Application layer commucate through Transport layer that to through a port. Ports are numbered and standard applications always use the same port. For example , SMTP protocol always use port 25, and HTTP protocol always use port 89 and FTP protocol always use ports 20 for data transmission and 21 for control.[6]

Transport layer:-

The layers gets data from the Application layer and divide into several data packets. It is the most uses TCP. After receiving data TCP protocol gets the packets which sent by the Internet layer and make the in an order, because arrives at the destiny in out of order. Transport layer also check if the content of the received packet are intact and reply by giving acknowledge signal to the transmitter , allowing it to know that the packet arrived in an order at the destiny. If there are no acknowledge signal is received it re send the lost packet.[6]

Internet layer: –

IP address is the unique virtual identity on TCP/IP network of each computer. It as charge of adding a header to the data packets received from the Transport layer. Where other control data will, it will add source IP address and target IP address. the physical address assigned to the network card of each computer. This address is written on the network card RO and is call MAC address. Let us suppose A computer want send a data to B computer so the A want to know the B MAC address but in small local area network computer can easily discover each other MAC address, this is an easy task. If there is no any virtual addressing were used we must to know the MAC address of the destination computer, which is not only a hard task but also does not help out packet routing, because it does not use a tree like structure.[6]

Network Interface Layer:-

The next is the network interface layer, the datagram generated on the upper layer will be sent to the network interface layer if we are sending data, or the Network Interface layer will get data from the network and send it to the Internet layer, if we are receiving data. The Network Interface Layer breaks down the packets from the Internet Layer into frames and then eventually into bits for transmission across the physical network medium. Signalling and network medium standards such as Ethernet, Token Ring, FDDI, X .25, Frame Relay, RS-232, and v.35 are defined in this layer. Network Interface Controller (NIC) Cards, network hubs, repeaters, bridges, and switches operate at this level.[6]

TCP/IP and OSI MODEL:-

TCP/IP is defined in term of the protocols that constitute it. Most of the critical protocols functions are their lower layer of the OSI Reference MODEL that are layer 2, 3 and 4 which correspond to the network interface that is internet and transport layer in the TCP/IP model. Which includes all important IP at layer 3 and TCP at layer 4 which combine to give TCP/IP the name to it. TCP contains four subsections. The first description are the

two TCP/IP protocols that reside at the network layer, layer 2 of the OSI model PPP and SLIP. And the second details a couple of special protocol that reside between layer 2 and 3 ARP and RARP. The third covers the TCP/IP internet layer (i.e) OSI network layer and layer 3 including IP and several other related and support protocol. At the final the fourth describes the TCP/IP transports layer protocols TCP and UDP.[7]

The lower layer of the OSI model is the physical layer which responsible for the transmitting information from one place to another on a network. The layer just above the physical layer is the data link layer and in TCP/IP its call has Network interface layer. Its main job is to implement network at local level and interface between the hardware oriented physical layer, and the ore abstract, software oriented function of the network layer and the above it.[8]

Effects of Internet:-

As a consequence of the development of internet, the internet protocol best effort delivery. But it is considered to be unreliable when it is talked about its service. In other network architectural language it is called as connection-less protocol. Thus it has its contrast implications to so called connecting modes of transmission. This lack of reliability causes many effects. The major problems raised are:

Data corruption

Lost data packets

Duplicate arrival

Out-of-order packet delivery

Coming to the effects based on IPv4 and IPv6, they are as follows.

IETF declared that IPv4 will not be able to accommodate the need for ever increasing amount of IP addresses in the next few years. The increase and expansion of IP addresses are needed to accommodate following. They are used for proliferation of internet devices such as personal computers, personal digital assistants (PDAs), wireless devices, and new Internet appliances; the expansion of internet through the world; the increasing applications of internet access; requirements of emerging internet applications. Thus decrease in usage of IPv4 effects all the above mentioned utilizes. The solution for this problem is found from development of IPv6 which enables a mass market and the increased adoption of the internet.

IPv4 was in existence and effectively working from almost 20 years. But now all of it’s approximately four billion of address seems to be getting finished. Because with the tremendous growth of the internet devices and blocks of the ip addresses assigned to various organizations and countries. This fast growth in the number of devices e.g. Mobiles, computers, music players, etc. it looked as if IPv4 addresses would be exhausted in near future. In order to overcome this problem and increasing life for IPv4 certain measures have been taken e.g. network address translation (NAT), classless inter-domain routing (CIDR) and PPP/DHCP. But all of these solutions are temporary and they made networks more complex and difficult for one to understand and implement.

Internet Stack Model

The Internet Stack Model employs a Internet Protocol Stack or Suite both these words are used interchangeably. This can be defined as the set of communications protocols used for the Internet and other similar networks. As discussed earlier Protocol is a set of rules in a network. The names of the two protocols which employed the Internet Stack Model are Transmission Control Protocol (TCP) and the Internet Protocol (IP), which were also the first two networking protocols defined in this standard. A protocol stack (sometimes communications stack) is a particular software implementation of a computer networking protocol suit. In practical implementation, protocol stacks are often divided into three major sections: media, transport, and applications. A particular operating system or platform will often have two well-defined software interfaces: one between the media and transport layers, and one between the transport layers and applications.

Upper layers are logically closer to the user and deal with more abstract data, relying on lower layer protocols to translate data into forms that can eventually be physically transmitted.[1]

Difference between OSI and TCP/IP:-

OSI model is the standard reference model that describes how the protocols should interact with one another. This was invented by the Department of Defense.TCP/IP does map clearly into OSI model, but is so convenient to think in terms of the OSI model when explaining protocols.

The major differences are:-

The application layer in TCP/IP handles the responsibilities of layers 5, 6 and 7 in the OSI model.

The transport layer in TCP/IP does not always guarantee reliable delivery of packets at the transport layer, while the OSI model does. TCP/IP also offers an option called UDP that does not guarantee reliable packet delivery. [9]

TCP/IP appears to be a more simpler model and this is mainly due to the fact that it has fewer layers

TCP/IP is considered to be 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 due to this reason. Where as in contrast networks are not usually built around the OSI model as it is merely used as a guidance tool

The OSI model consists of 7 architectural layers whereas the TCP/IP only has 4 layers.

Disadvantages of OSI model:

Major disadvantages of OSI model are

Bad timing

Bad technology

Bad implementations

Bad Timing: –

Presently it appears that standard OSI protocols got crushed. It was completed too late, TCP/IP had already taken hold and become the familiar standard for research universities by the time the OSI Model was ready

Bad Technology:-

Both the models and protocols are flawed. As mentioned about the layers, cause major disadvantage of this model. Working on OSI model which is associated with several definitions and protocols is extraordinarily complex. The reappearance of functions in each layer also cause problem. Some of those functions are addressing, flow control, and error control. Certain important features are placed in every layer. Some times the decisions taken for placing these features are not obvious. In the past the presentation layer had virtual terminal handling which is presently used in application layer. Data security, encryption and network management were also omitted. The criticism given by world is that communication mentality is increasing rapidly.

Bad Implementations: –

Initial implementations made by people were huge ,unwieldy and slow Because of the complexity of the model, the initial implementations were huge, unwieldy, and slow. As time is passing the product got betterments, but bad image stuck. [10]

everyone who tried to implement it got burned, giving OSI a reputation for poor quality

Disadvantages of TCP/IP model:

When TCP/IP is compared with IPX there are few drawbacks.

IPX is faster compared to TCP/IP.

TCP/IP is intricate to set up and manage.

The overhead of TCP/IP is higher than that of IPX.

The process involved in TCP/IP is long. The entire process is, port numbers are assigned by the Internet Assigned Numbers Authority (IANA), and they identify the process to which a particular packet is connected to. Port numbers are found in the packet header. [11].

Few more disadvantages of this model are: software used for strict layering is very inefficient. When buffers are provided they leave space for headers which are added by lower layer protocols. [12]

Security:-

Internet has become a phenomenon. Sharing Files, transferring Files all across the network has caused the users to think about some serious problems that are being raised due to the phenomenon of Internet. Security is another issue that sits at the top of the level as designers and developers both struggle to give out the correct form of measures to be taken to deal with the issue of Network Security. Since the Protocol Stack has got just one layer to really check the authenticity of the packets that pass through the layers and the network. It has allowed the network to become more susceptible to foreign threats. Since, Computers that are connected to each other create a network. These networks are often configured with “public” Internet Protocol (IP) addresses — that is, the devices on the network are “visible” to devices outside the network. Networks can also be configured as “private” — meaning that devices outside the network cannot “see” or communicate directly to them.

Computers on a public network have the advantage (and disadvantage) that they are completely visible to the Internet. As such, they have no boundaries between themselves and the rest of the Internet community. This advantage oftentimes becomes a distinct disadvantage since this visibility can lead to a computer vulnerability exploit -e.g.: Hacking — if the devices on the public network are not properly secured.
To plainly counter this strategy the, we employ a private LAN and make the network a private network then the disadvantage of a private network would be that it entails more configuration and administration to maintain usability. At times, not being fully visible on the Internet can cause some difficulty in connecting to certain services, such as streaming audio/video, chat/instant messaging programs, or some secure Web sites.

Maintaining most computers on a private network, with only an IDP/IDS and/or Firewall visible to the public Internet helps maintain a highly secure environment for the computers on the private network, while at the same time keeping them connected to the public Internet.[13]

Benefits of OSI model:-

The OSI Model is a standard and a very useful tool to deal with the problems of networking. The OSI model is also a standard model for networking protocols and distributed application. In the OSI model there are a total of 7 network layers. The layers of OSI provide the levels of abstraction. Each layer performs a different set of functions and the intent was to make each layer as independent as possible from all the others. This guarantees the security to each packet that passes through these layers also, it reduces the complexity of packet routing through each layer thus reducing the threat of congestion in the network.

Each layer uses the information from the below layer and provides a service to the layer above in the OSI Model. [14]

Conclusion:

Even though OSI model has a few limitations it cannot be said that the model is not suitable anymore because talking about the time and the influence of the OSI model on the networking as whole has had a lot of positive causes. Even if we have many new models coming up with different improvements the base model is always the OSI model. What we can take out of the OSI model are mainly the benefits of the OSI model and which even are the major characteristics needed for creating a good model. The OSI Model is perhaps the most imp lest and widely used through the world. Owing to the advantages of this network we are able to reduce the complexity and save time when passing data on a network.

Reference:-

1) John Larmouth (1996): Understanding OSI: International Thomson Computer Press, Michigan. USA}

2) Computer networks 3rd edition by ANDREW S. TANENBAUM

3) Standard network layer architecture, By Bradley Mitchell http://compnetworking.about.com/cs/designosimodel/a/osimodel.htm

4)http://www2.rad.com/networks/introductory/layers/main.htm

5) By Douglas Comer’s http://www.cellsoft.de/telecom/tcpip.htm

6) Author: Gabriel Torres http://www.hardwaresecrets.com/article/433/1,2,3,4,5,6

7) Charles M. Kozierok

.http://www.tcpipguide.com/free/t_TCPIPLowerLayerInterfaceInternetandTransportProtoc.htm

8) Charles M. Kozierok

http://www.tcpipguide.com/free/t_DataLinkLayerTechnologiesandProtocols.htm

http://itprofesionals.blogspot.com/2009/01/what-is-difference-between-osi-model.html

10) Computer Networks Fourth Edition by Andrew S. Tanenbaum

11)www.tech-faq.com/understanding-the-osi-model.shtml

www.bookrags.com/wiki/OSI_model

13) Behrouz A. Forouzan, Sophia Chung Fegan (2004). Data Communications and Networking, McGraw-Hill Professional, USA.

14) http://compnetworking.about.com/cs/designosimodel/a/osimodel.htm