By Kailin Acheson
The previous two parts of this blog, Part I and Part II, covered the seven layers of the Open Systems Interconnection (OSI) model. This installment will cover the four layers of the Transmission Control Protocol/Internet Protocol (TCP/IP) model and contrast the TCP/IP model with the OSI model.
The TCP/IP model and the OSI model are related in that they are layered reference models that represent how data and information are passed between various devices. But they are different in that the TCP/IP model has four layers (Application, Transport, Internet, and Network Access) and the OSI model has seven layers (Physical, Data Link, Network, Transport, Session, Presentation, and Application). The following graphic shows how the layers relate to one another when compared side by side:
As you can see, the TCP/IP model has fewer layers, which in some ways overlap the layers of the OSI model. Whereas the OSI model emphasizes certain layers and was structured with the layers as the primary focus, the TCP/IP model was structured with architectural principles as the primary focus.
To refresh your memory on the layers of the OSI model, check out Part I (which covers the Physical, Data Link, Network, and Transport layers) and Part II (which covers the Session, Presentation, and Application layers). I won't go into an exhaustive discussion of each layer; what you see below will be a brief summary of each layer so you have an idea of the basics.
The TCP/IP Application layer, also called the Process layer, is the top layer, closest to you, the user, and is responsible for delivering data and network services to the specific applications on a computer for which they are intended. The Application layer is also responsible for the compression and encryption of data as well as the creation, maintenance, and termination of sessions when required. Some common protocols that operate at the TCP/IP Application layer include File Transfer Protocol (FTP), Simple Mail Transfer Protocol (SMTP), Domain Name System (DNS), Telnet, and Hypertext Transfer Protocol (HTTP). The Application layer of the TCP/IP model corresponds to the Application layer (Layer 7), the Presentation layer (Layer 6), and the Session layer (Layer 5) of the OSI model.
The TCP/IP Transport layer, also called the Host-to-Host layer, is responsible for the delivery of data between the Internet layer and the Application layer, error control and recovery, flow control, congestion control, and data segmentation. The common protocols that operate at the TCP/IP Transport layer are Transmission Control Protocol (TCP) and User Datagram Protocol (UDP). The devices that typically operate at the Transport layer are network devices, or gateways. The Transport layer of the TCP/IP model corresponds to the Transport layer (Layer 4) of the OSI model.
The TCP/IP Internet layer, also known as the Internetwork layer, is responsible for defining logical addressing and delivering packets. The main protocol used at this layer, Internet Protocol, uses the source and destination addresses in order to facilitate the movement of data between the Network Access layer and the Transport layer; the data flows from one node on the network to the next node in a path moving toward the final destination. The two other common protocols that operate at the TCP/IP Internet layer are derivatives of IP: IP version 4 (IPv4) and IPv6. The devices that typically operate at the Internet layer are routers. The Internet layer of the TCP/IP model corresponds to the Network layer (Layer 3) of the OSI model.
The TCP/IP Network Access layer, also referred to as the Link layer or Network Interface layer, defines how data will be delivered over the physical network and which protocols are appropriate for that delivery. This layer is also responsible for arbitration and error discovery. The Network Access layer can at times be separated into two different layers: the Data Link layer and the Physical layer. Some common protocols that operate at the TCP/IP Network Access layer include Address Resolution Protocol (ARP) and Point-to-Point Protocol (PPP). The devices that operate at the Network Access layer include Layer 2 switches and bridges (operating on the Data Link layer) and Layer 1 hubs and repeaters (operating on the Physical layer). The Network Access layer of the TCP/IP model corresponds to the Data Link layer (Layer 2) and the Physical layer (Layer 1) of the OSI model.
Now that you have read the basics of each of the four TCP/IP layers, let's take a brief look at how they interact and how data moves between them. Much like the OSI model, a layer of the TCP/IP model can communicate with the layer directly above it or the layer directly below it (inter-area communication) or with the layer directly across from it on a connected device (intra-area communication). The devices that operate at each layer are noted below:
A hub operates at the Physical layer division of the Network Access layer. A hub, which cannot make forwarding decisions regarding the data it receives, simply receives, amplifies, and regenerates signals between ports. A hub can only receive a signal containing a sequence of bits and then transmit the same bits on all its other ports to a Layer 2 switch or bridge. Because the devices at this layer can only receive and retransmit data, no Protocol Data Unit (PDU), such as a packet or a segment, exists at this layer.
A Layer 2 switch or bridge operates at the Data Link layer division of the Network Access layer. The stream of bits a Layer 2 devices receives from the Physical layer becomes known as a frame PDU at the Data Link layer. A Layer 2 switch or bridge can receive a stream of bits from the Physical layer, verify that the entire frame has arrived safely, and check the integrity of the data by performing a calculation on the received data and comparing the result to the Frame Check Sequence (FCS). If the frame is verified, the Layer 2 device can forward the frame to the appropriate destination based on the Media Access Control (MAC) address and the information in the Content Addressable Memory (CAM) table.
A router operates at the Internet layer. The frame a router receives from the Network Access layer becomes known as a packet PDU at the Internet layer. The router can verify that the IP header information within the packet has remained intact and can determine whether the packet has been fragmented during the delivery process. The router can then make a forwarding decision based on the destination IP address in a packet's header and the entries in the router's own routing table.
Gateways operate at or above the Transport layer; an example of a gateway is a voice gateway that can accept an ISDN-PRI trunk interface that uses Q.931 from the phone company and translate it to the H.323 signaling protocol used by Voice over IP (VoIP) telephone systems. The packet a gateway receives from the Internet layer becomes known as a flow PDU at the Transport layer. A gateway receives a packet from the Internet layer and then makes forwarding decisions with more information than the routers at the Internet layer are able to use; for example, a gateway can use information from the Internet layer headers, higher layer headers, or application data.
Hosts operate at the Application layer. A host (which could be a desktop PC, notebook, laptop, wireless access point [WAP], or network-attached printer, among other devices) is typically the final destination and originator of a flow of data. A host does not forward packets, but it maintains a routing table to facilitate forwarding decisions for packets that originate from the host.
Data that flows between layers can be viewed from two different perspectives: inter-layer communication and intra-layer communication. Inter-layer communication is the transfer of information between adjacent TCP/IP layers on the same device and is represented by the vertical arrows in the graphic below; intra-layer communication is the transfer of information between identical TCP/IP layers on different devices and is represented by the horizontal arrows in the graphic below:
The discussions in all three parts of this blog are basic summaries of each of the model's different layers. They are designed to help you understand what others are talking about in a conversation about networking and to help you comprehend what someone means when he or she says that Device X operates at Layer Y, or that this problem is a Layer Z issue.
If you would like to learn more about the OSI model, Cisco's Open System Interconnection Protocols page is a good resource that provides a lot of additional detail. If you would like to learn more about the TCP/IP model, Wikipedia's Internet protocol suite page, Cisco's TCP/IP Overview page, and Request for Comments (RFC) 1122 all contain valuable information. Another resource you'll want to know about is our Cisco Courseware from Boson. In addition to more complete coverage of the OSI model than both John and I were able to provide in these blogs, this courseware covers all of the ICND1 and ICND2 topics you'll need to master as you prepare for your CCNA certification.