OSI model vs TCP/IP model | difference between tcp ip and osi layers
The OSI (Open Systems Interconnection) model and the TCP/IP (Transmission Control Protocol/Internet Protocol) model both describe the functions of network protocols and services required for communication between devices on a network. While the two models have different layering schemes, they share similar functions. OSI model consists of seven layers viz. application, presentation, session, transport, network, data link and physical. Initially, TCP IP model consists of four layers viz. application, transport, internet and network access. Later, it has been represented with five layers for easy comparison with OSI model in which "network access layer" has been represented by data link and physical layers.
The OSI(Open system Interconnection) is developed for defining seven layers which makes two different vendors systems communicate. TCP IP is used for internet applications mainly. The figure describes how TCP IP layers map to OSI layers.
Figure-1 : OSI layers vs TCP IP model layers
In OSI layers, control is passed from one layer to the next, starting at the Layer 7, application layer in one station, proceeding to the succeeding bottom layer and so on till information reaches Layer 1.Then the information transmitted over the channel to the next station's Layer 1 and transmitted to succeeding layers till information delivered to Layer 7. Each Layer in the OSI model does the job for the previous layer and also adds prefixes information termed as "header". The data link layer adds suffixes called cyclic redundancy check as "trailer" which is used for error detection. The brief functionality of each OSI layer is mentioned below. Let us understand functions of OSI layers and TCP IP layers.
Functions of OSI and TCP IP layers
Application layer :
All the applications for user reside at this layer. It includes FTP, Telnet, e-mail etc.
It formats and encrypts data to be sent across a network. Its basic task is to make application layer and network layer talk to each other by processing the data. It provides independence to the application processes from differences in data representation by way of syntax conversion.
Session layer :
It Provides control mechanism to connect two applications running on different machines. It covers basic function to establish, to maintain and to release connections.
Transport layer :
It assures reliable data communication between two hosts. It is responsible for error recovery and flow control between two hosts. It also does multiplexing and de multiplexing functionalities for transmitting and receiving host respectively. The tcp and udp reside at transport layer of the osi model.
Network layer :
The main function is switching and routing of the packet. It adds logical addressing to do this. It also takes care of error handling, congestion control and packet sequencing.
Data Link layer :
It takes care of errors in the physical layer. It provides flow control and also does frame synchronization. ARQ is implemented to take care of retransmission in case of erroneous packets. In latest today's technologies it provides Medium Access Control functionality, hence known as MAC. MAC takes care of resource assignment to the end devices wanted to seek connection to the network for various applications (voice, data, etc.)
Physical layer :
It mainly takes care of bit stream transmission over physical medium (cable, air etc.). It provides the hardware means of sending and receiving data through cable or wireless medium. Some wireless standards viz. wlan, wimax, zigbee, LTE defines physical layer specifications which takes care of forward error correction and modulation-demodulation functionalities as well as making the information suitable to be transmitted. Following figure mentions logical structure of the layered protocols inside a computer on an internet. This figure is taken from RFC 1180 to describe the concept.
RFC1180 describes logical structure of protocols inside a typical internet enabled
device as mentioned below.
Ethernet Cable--ENET Card--ARP/IP--TCP/UDP--Network Applications
As mentioned etnernet cable is connected directly with Ethernet card which has upper layer ARP or IP. This has upper layer TCP or UDP and above all applications are running.
Every device has its own Ethernet address which is 6 bytes and listens for Ethernet frames with that destination address. An IP address is 4 bytes. Ethernet uses CSMA/CD. CSMA/CD means that all devices communicate on a single medium in which only one can transmit at a time and they can all receive simultaneously.
If two devices try to transmit at the same instant, the transmit collision is detected and both devices wait a random period of time before trying to transmit again. An Ethernet frame contains the destination address, source address, type field, and data.
Data flow across TCP IP model layers
Let us understand data flow communication between TCP IP model layers of Host-A and Host-B.
Let us assume host A is transmitting packet and Host B is receiving it.
• As mentioned above all the hosts receive the packets and each will compare destination address in the Ethernet header with its' own Ethernet or MAC address of 6 bytes and accept the same if it matches otherwise does not pay attention.
• Next it will check for type field(2 byte long), if it is 0x806 then the packet will be passed to ARP module and if it is 0x800 it will be passed to IP module. ARP module responds to hosts seeking for Ethernet address based on their logical (IP) address.
• After packet is received by IP layer it will decide where the packet need to be routed based on routing table and destination IP address embedded in received packet's IP header. IP header is usually 20 bytes.
• If the host is the final end system then IP module checks for protocol field (1 byte long) in the IP header. If it is 0x06 then the packet is passed to the TCP module and if it is 0x11 then the packet is passed to the UDP module.
• Above the TCP or UDP there is application layer, where so many applications will be usually running. Information to each application will be passed based on 'destination port number' field (16 bit long) embedded inside the TCP or UDP header. TCP utilizes 'sequence number field' (32 bit long) for reassembly of all the IP data grams received at irregular time instants in irregular orders. For example FTP has dedicated port address 23 and Telnet has dedicated port address 21, which is used to deliver the data to each.
Comparison between OSI layers vs TCP IP layers
Let us compare osi model vs tcp/ip model and understand difference between tcp ip and osi model layers with their functions and other comparison parameters.
Following are the benefits or advantages of OSI model.
➨OSI model provides clear and structured conceptual framework to understand network communication.
➨OSI model's modularity allows implementation of protocols for specific functionalities.
➨It promotes interoperability between equipments from different manufactures or vendors.
➨As an ISO standard it helps to establish uniformity in network design and encourages to follow common standards.
➨It's layered nature makes it easy to design, troubleshoot and upgrade individual layers without affecting others.
➨OSI model remains fundamental tool to learn basics of networking.
Following are the benefits or advantages of TCP/IP model.
➨It has fewer layers and streamlined structure compare to OSI model which makes it easy to implement and troubleshoot in real world networking scenarios.
➨It is fundamental networking protocol suite used on internet. It is widely adopted globally to connect diverse networks.
➨TCP/IP model's flexible design allows it to scale as per increasing demands of network resources. New devices and technologies are easily incorporated without significant modifications.
➨It's direct use in practical networking including internet helps network administrators, engineers and IT professionals working in real world networks.
➨TCP IP has been in use since several decades which has made it reliable and robust. It has become stable choice for network communication.
Conclusion : Overall, both tcp ip and osi models provide a framework for understanding the different functions involved in network communication, from user applications down to the physical transmission of data. The OSI model is more comprehensive and theoretical, while the TCP/IP model is more streamlined and practical.
TCP IP protocol stack development community
• Internet Architecture Board (IAB), www.iab.org/
• Internet Engineering Task Force (IETF), www.ietf.org/
• Internet Society, www.isoc.org/
• The Internet Corporation of Assigned Names and Numbers (ICANN), www.icann.org/
• Internet Research Task Force (IRTF), www.irtf.org/
• World Wide Web Consortium (W3C), www.w3.org/
• RFC 1180, A TCP IP Tutorial
• RFC793, TCP
• RFC826, ARP