OSI Model Overview

The International Organization for Standardization (ISO) developed the OSI model in 1984, and revisited it in 1994, to coordinate standards development for interconnected information-processing systems. The model describes seven layers that start with the physical connection and end with the application. As shown in Figure C-1, the seven layers are physical, data link, network, transport, session, presentation, and application.

Physical Layer (OSI Layer 1)

The physical layer describes the transportation of raw bits over physical media. It defines signaling specifications and media types and interfaces. It also describes voltage levels, physical data rates, and maximum transmission distances. In summary, it deals with the electrical, mechanical, functional, and procedural specifications for links between networked systems.

Data Link Layer (OSI Layer 2)

The data link layer is concerned with the reliable transport of data across a physical link. Data at this layer is formatted into frames. Data link specifications include frame sequencing, flow control, synchronization, error notification, physical network topology, and physical addressing. This layer converts frames into bits when sending information and converts bits into frames when receiving information from the physical media. Bridges and switches operate at the data link layer.

Network Layer (OSI Layer 3)

The network layer is concerned with routing information and methods to determine paths to a destination. Information at this layer is called packets. Specifications include routing protocols, logical network addressing, and packet fragmentation. Routers operate at this layer.

Transport Layer (OSI Layer 4)

The transport layer provides reliable, transparent transport of data segments from upper layers. It provides end-to-end error checking and recovery, multiplexing, virtual circuit management, and flow control. Messages are assigned a sequence number at the transmission end. At the receiving end, the packets are reassembled, checked for errors, and acknowledged. Flow control manages the data transmission to ensure that the transmitting device does not send more data than the receiving device can process.

Session Layer (OSI Layer 5)

The session layer provides a control structure for communication between applications. It establishes, manages, and terminates communication connections called sessions. Communication sessions consist of service requests and responses that occur between applications on different devices.

Presentation Layer (OSI Layer 6)

The presentation layer provides application layer entities with services to ensure that information is preserved during transfer. Knowledge of the syntax selected at the application layer allows selection of compatible transfer syntax if a change is required. This layer provides conversion of character-representation formats, as might be required for reliable transfer. Voice coding schemes are specified at this layer. Furthermore, compression and encryption can occur at this layer.

Application Layer (OSI Layer 7)

The application layer gives the user or operating system access to the network services. It interacts with software applications by identifying communication resources, determining network availability, and distributing information services. It also provides synchronization between the peer applications residing on separate systems.

TCP/IP Architecture

The suite of TCP/IP protocols was developed for use by the U.S. government and research universities. The suite is identified by its most widely known protocols: TCP and IP. As mentioned, the ISO published the OSI model in 1984. However, the TCP/IP protocols had been developed by the Department of Defense’s Advanced Research Projects Agency (DARPA) since 1969. The TCP/IP uses only four layers (as described in RFC 791) versus the seven layers used by OSI. The TCP/IP layers are

Network Interface Layer

The TCP/IP network interface layer (also known as network access layer) maps to the OSI data link and physical layers. TCP/IP uses the lower-layer protocols for transport.

Internet Layer

The Internet layer is where IP resides. IP packets exist at this layer. It directly maps to the network layer of the OSI model. Other TCP/IP protocols at this layer are Internet Control Message Protocol (ICMP), Address Resolution Protocol (ARP), and Reverse ARP (RARP).

Host-to-Host Transport Layer

The host-to-host transport layer of TCP/IP provides two connection services: TCP and UDP. TCP provides reliable transport of IP packets, and UDP provides transport of IP packets without verification of delivery. This layer maps to the OSI transport layer, but the OSI model only defines reliable delivery at this layer.

Application Layer

The TCP/IP application layer maps to the top three layers of the OSI model: application, presentation, and session. This layer interfaces with the end user and provides for authentication, compression, and formatting. The application protocol determines the data’s format and how the session is controlled. Examples of TCP/IP application protocols are Telnet, FTP, BGP, and Hypertext Transfer Protocol Secure (HTTPS).

Example of Layered Communication

Suppose that you use a Telnet application. Telnet maps to the top three layers of the OSI model. In Figure C-4, a user on Host 1 enables the Telnet application to access a remote host (Host 2). The Telnet application provides a user interface (application layer) to network services. As defined in RFC 854, ASCII is the default code format. No session layer is defined for Telnet (not an OSI protocol). Per the RFC, Telnet uses TCP for connectivity (transport layer). The TCP segment is placed in an IP packet (network layer) with a destination IP address of Host 2. The IP packet is placed in an Ethernet frame (data link layer), which is converted into bits and sent onto the wire (physical layer).

Numeric Conversion

This section focuses on the techniques for converting between decimal, binary, and hexadecimal numbers. Although the exam might not have a specific question about converting a binary number to decimal, you need to know how to convert these numbers to do problems on the test. An IPv4 address could be shown as binary or in traditional dotted-decimal format. MAC addresses and IPv6 addresses are represented in hexadecimal. Some show commands have output information in hexadecimal or binary formats.

Hexadecimal Numbers

The hexadecimal numeric system uses 16 digits instead of the 10 digits used by the decimal system. Table C-1 shows the hexadecimal digits and their decimal equivalent values.

Hexadecimal Representation

It is common to represent a hexadecimal number with 0x before the number so that it is not confused with a decimal number. The hexadecimal number of decimal 16 is written as 0x10, not 10. Another method is to put a subscript h to the right of the number, such as 10_h. It is also common to use the term hex when speaking of hexadecimal. Much of the following text uses hex.

Converting Decimal to Hexadecimal

First things first: Memorize Table C-1. There are two ways to convert larger numbers. The first method is to convert decimal to binary and then convert binary to hex. The second method is to divide the decimal number by 16—the residual is the rightmost hexadecimal digit—and then keep dividing until the number is not divisible anymore. For the first method, use the schemes described in later sections. For the second method, follow the examples described here.

Converting Hexadecimal to Decimal

To convert a hex number to decimal, take the rightmost digit and convert it to decimal (for example, 0xC = 12). Then add this number to the second rightmost digit multiplied by 16 and the third rightmost digit multiplied by 256. Do not expect to convert numbers larger than 255 on the CCDA exam, because the upper limit of IP addresses in dotted-decimal format is 255 (although Token Ring numbers reach 4096). Some examples follow.

Alternative Method for Converting from Hexadecimal to Decimal

Another way is to convert from hex to binary and then from binary to decimal. The following sections discuss converting from binary to decimal.

Binary Numbers

The binary number system uses two digits: 1 and 0. Computer systems use binary numbers. IP addresses and MAC addresses are represented by binary numbers. The number of binary 1s or 0s is the number of bits, short for binary digits. For example, 01101010 is a binary number with 8 bits. An IP address has 32 bits, and a MAC address has 48 bits. As shown in Table C-2, IPv4 addresses are usually represented in dotted-decimal format; therefore, it is helpful to know how to convert between binary and decimal numbers.

Converting Binary to Hexadecimal

To convert binary numbers to hex, put the bits in groups of 4, starting with the right-justified bits. Groups of 4 bits are often called nibbles. Each nibble can be represented by a single hexadecimal digit. A group of two nibbles is an octet, 8 bits. Examples follow.

Converting Hexadecimal to Binary

This procedure is also easy. Just change the hex digits into their 4-bit equivalents. Examples follow.

Converting Binary to Decimal

To convert a binary number to decimal, multiply each instance of 0 or 1 by the power of 2 associated with the position of the bit in the binary number. The first bit, starting from the right, is associated with 2⁰ = 1. The value of the exponent increases by 1 as each bit is processed, working leftward. As shown in Table C-4, each bit in the binary number 10101010 has a decimal equivalent from 0 to 128 based on the value of the bit multiplied by a power of 2 associated with the bit position. This is similar to decimal numbers, in which the numbers are based on powers of 10: 1s, 10s, 100s, and so on. In decimal, the number 111 is (1*100) + (1*10) + (1*1). In binary, the number 11111111 is the sum of (1*2⁷) + (1*2⁶) + (1*2⁵) + (1*2⁴) + (1*2³) + (1*2²) + (1*2¹) + (1*2⁰) = 128 + 64 + 32 + 16 + 8 + 4 + 2 + 1 = 255. For 10101010, the result is 128 + 0 + 32 + 0 + 8 + 0 + 2 + 0 = 170. Examples follow.

Converting Decimal to Binary Numbers

This procedure is similar to converting from hex to decimal (by dividing), but now you divide the decimal number by 2. You use each residual to build the binary number by prepending each residual bit to the previous bit, starting on the right. Repeat the procedure until you cannot divide anymore. The only problem is that for large numbers, you might have to divide many times. You can reduce the number of divisions by first converting the decimal value to a hexadecimal value and then converting the intermediate result to the binary representation. After the following example, you will read about an alternate method suitable for use with decimal values between 0 and 255 that can be represented in a single octet.

Alternative Method for Converting from Decimal to Binary

The dividing procedure just described works; it just takes a lot of time. Another way is to remember the bit position values within a byte—128, 64, 32, 16, 8, 4, 2, 1—and play with the bits until the sum adds up to the desired number. This method works when you convert integer values between 0 and 255, inclusive. Table C-5 shows these binary numbers and their decimal values.

References and Recommended Readings

ISO/IEC 7498-1: 1994. “Information Processing Systems - OSI Reference Model - The Basic Model.”