Chapter 2: Networking Standards Organizations and the OSI Model

On TCP/IP-based networks

(such as the Internet), Network layer protocols can perform an additional function called fragmentation.

IAB (Internet Architecture Board)

IAB is a technical advisory group of researchers and technical professionals interested in overseeing the Internet's design and management. IAB is responsible for Internet growth and management strategy, resolution of technical disputes, and standards oversight.

In the example of requesting a Web page

IP is the protocol that instructs the network where the HTTP request is coming from and where it should go. Figure 2-7 depicts the data found in an IP packet used to contact the Web site www.loc.gov/index.html. Notice the Network layer addresses, or IP addresses, in the first line of the packet. The first, labeled "src Addr" reveals the unique IP address of the com- puter issuing the transmission. The next, labeled "DST Add," indicates the unique IP address of the receiving computer.

Session Layer Connection

If a connection is lost, the Session layer protocols will detect that and initiate attempts to reconnect. If they cannot reconnect after a certain period of time, they will close the session and inform your client software that communication has ended. Finally, the Session layer monitors the identification of session participants, ensuring that only the authorized nodes can access the session.

For example

If you look up your bank account status via the Internet, you are using a secure connection, and Presentation layer protocols will encrypt your account data before it is transmitted. On your end of the network, the Presentation layer will decrypt the data as it is received.

TIA (Telecommunications Industry Association)

In 1988, one of the EIA's subgroups merged with the former United States Telecommunications Suppliers Association (USTSA) to form TIA. TIA focuses on standards for information technology, wireless, satellite, fiber optics, and telephone equipment. Both TIA and EIA set standards, lobby governments and industry, and sponsor conferences, exhibitions, and forums in their areas of interest.

Segmentation

In addition to ensuring reliable data delivery, Transport layer protocols break large data units received from the Session layer into multiple smaller units, called segments.

OUI (Organizationally Unique Identifier)

Is the first part of the Physical Address. It is a character sequence assigned by IEEE that identifies the NIC's manufacturer.

In the example of retrieving a Web page

A Transport layer protocol called TCP (Transmission Control Protocol) takes care of reliably transmitting the HTTP protocol's request from client to server and vice versa.

Each node has two types of addresses.

Network addresses follow a hierarchical addressing scheme and can be assigned through operating system software. They are hierarchical because they contain subsets of data that incrementally narrow down the location of a node, just as your home address is hierarchical because it provides a country, state, zip code, city, street, house number, and person's name. Network layer address formats differ depending on which Network layer protocol the network uses. Network addresses are also called Network layer addresses, logical addresses, or virtual addresses. The second type of address assigned to each node is called a physical address, discussed in detail in the next section.

Application Layer

The Application layer facilitates communication between such programs (Microsoft Word, Firefox ect.) and lower-layer network services. Services at this layer enable the network to interpret a program's request and the program to interpret data sent from the network. Through Application layer protocols, programs negotiate their formatting, procedural, security, synchronization, and other requirements with the network. Note that not all these requirements are fulfilled by Application layer protocols. They are merely agreed upon at this stage.

FCS (frame check sequence)

The field in a frame responsible for ensuring that data carried by the frame arrives intact. It uses an algorithm, such as CRC, to accomplish this verification.

MAC (Media Access Control) sublayer

The lower sublayer of the Data Link layer, manages access to the physical medium. It appends the physical address of the destination computer onto the data frame.

Most common Network layer protocol

The most common, and the one that underlies most Internet traffic, is the IP (Internet Protocol).

SYN-ACK (synchronization-acknowledgment)

The packet a node sends to acknowledge to another node that it has received a SYN request for connection. The SYN-ACK packet is the second of three in the three-step process of establishing a connection.

SYN (synchronization)

The packet one node sends to request a connection with another node on the network. The SYN packet is the first of three in the three-step process of establishing a connection.

Network Layer

The primary function of protocols is to translate network addresses into their physical counterparts and decide how to route data from the sender to the receiver. Addressing is a system for assigning unique identification numbers to devices on a network.

For example

when you choose to open a Web page in Firefox, an Application layer protocol called HTTP (Hypertext Transfer Protocol) formats and sends your request from your client's browser (a software application) to the server. It also formats and sends the Web server's response back to your client's browser.

Example

What happens if the room is so noisy that Ms. Jones hears only part of your question? For example, she might receive "on commerce in the late-nineteenth century?" This kind of error can happen in network communications as well (because of wiring problems, for example). The Data Link layer protocols find out that information has been dropped and ask the first computer to retransmit its message—just as in a classroom setting Ms. Jones might say, "I didn't hear you. Can you repeat the question?" The Data Link layer accomplishes this task through a process called error checking.

CRC (cyclic redundancy check)

When the source node transmits the data, it performs an algorithm (or mathematical routine) called a _____. CRC takes the values of all of the preceding fields in the frame and generates a unique 4-byte number, the FCS.

NIC's physical address

You can find a NIC's physical address through your computer's protocol configuration utility or by simply looking at the NIC. The physical address will be stamped directly onto the NIC's circuit board or on a sticker attached to some part of the NIC.

Some Transport layer protocols take steps..

...steps to ensure that data arrives exactly as it was sent. Such protocols are connection oriented because they establish a connection with another node before they begin transmitting data. TCP is one example of a connection-oriented protocol. In the case of requesting a Web page, the client's TCP protocol first sends a SYN (synchronization) packet request for a connection to the Web server. The Web server responds with a SYN-ACK (synchronization-acknowledgment) packet, or a confirmation, to indicate that it's willing to make a connection. Then, the client responds with its own ACK (acknowledgment). Through this three-step process, also known as a three-way handshake, a connection is established. Only after TCP establishes this connection does it transmit the HTTP request for a Web page.

Connectionless Protocol Efficiency

A connectionless protocol's lack of sophistication makes it more efficient than a connection-oriented protocol and renders it useful in situations in which data must be transferred quickly, such as live audio or video transmissions over the Internet. In these cases, connection-oriented protocols -with their acknowledgments, checksums, and flow control mechanisms- would add overhead to the transmission and potentially bog it down. In a video transmission, for example, this could result in pictures that are incomplete or aren't updated quickly enough to coincide with the audio.

Terminal

A device with little (if any) of its own processing or disk capacity that depends on a host to supply it with applications and data-processing services.

Standard

A documented agreement containing technical specifications or other precise criteria that are used as guidelines to ensure that materials, products, processes, and services suit their intended purpose.

Flow Control

A method of gauging the appropriate rate of data transmission based on how fast the recipient can accept data.

Token Ring

A much less-common networking technology, was developed by IBM in the 1980s. It relies upon direct links between nodes and a ring topology. Nodes pass around tokens, special control frames that indicate to the network when a particular node is about to transmit data. Although this networking technology is nearly obsolete, there is a remote chance that you might work on a token ring network. The IEEE has defined token ring technology in its 802.5 standard.

Three-Way Handshake

A three-step process in which Transport layer protocols establish a connection between nodes. The three steps are: Node A issues a SYN packet to node B, node B responds with SYN-ACK, and node A responds with ACK.

Connection Oriented

A type of Transport layer protocol that requires the establishment of a connection between communicating nodes before it will transmit data.

PDUs (protocol data units)

A unit of data at any layer of the OSI model.

IP (Internet Protocol) addresses

Addresses used to identify computers on the Internet and other TCP/IP-based networks.

Programmers Dare Not Throw Salty Pretzels Away.

Networking professionals often devise a mnemonic way of remembering the seven layers of the OSI model. One strategy is to make a sentence using words that begin with the same first letter of each layer, starting with either the lowest (Physical) or the highest (Application) layer.

When receiving data

Physical layer protocols detect and accept signals, which they pass on to the Data Link layer. Physical layer protocols also set the data transmission rate and monitor data error rates. However, even if they recognize an error, they cannot perform error correction. When you install a NIC in your desktop PC and connect it to a cable, you are establishing the foundation that allows the computer to be networked. In other words, you are providing a Physical layer.

For Example

Pretend for a moment that computers communicate as humans do. Suppose you are in Ms. Jones's large classroom, which is full of noisy students, and you need to ask the teacher a question. To get your message through, you might say, "Ms. Jones? Can you explain more about the effects of railroads on commerce in the mid- nineteenth century?" In this example, you are the sender (in a busy network) and you have addressed your recipient, Ms. Jones, just as the Data Link layer addresses another computer on the network. In addition, you have formatted your thought as a question, just as the Data Link layer formats data into frames that can be interpreted by receiving computers.

Presentation Layer

Protocols at this layer accept Application layer data and format it so that one type of application and host can understand data from another type of application and host. In other words, the Presentation layer serves as a translator. Presentation layer services also manage data encryption (such as the scrambling of passwords) and decryption.

Transport Layer

Protocols in this layer accept data from the Session layer and manage end-to-end delivery of data. That means they can ensure that the data are transferred from point A to point B reliably, in the correct sequence, and without errors. Without Transport layer services, data could not be verified or interpreted by its recipient. Transport layer protocols also handle flow control, which is the process of gauging the appropriate rate of transmission based on how fast the recipient can accept data.

Session Layer

Protocols in this layer coordinate and maintain communications between two nodes on the network. Functions are establishing and keeping alive the communications link for the duration of the session, keeping the communication secure, synchronizing the dialogue between the two nodes, determining whether communications have been cut off, and, if so, figuring out where to restart transmission, and terminating communications. Session layer services also set the terms of communication by deciding which node communicates first and how long a node can communicate.

Segmentation Process

Segmentation is necessary for data units to match a network's MTU (maximum transmission unit), the largest data unit it will carry. Every network type specifies a default MTU (though its size can be modified to some extent by a network administrator). For example, by default, Ethernet networks cannot accept packets with data payloads larger than 1500 bytes. Suppose an application wants to send a 6000-byte unit of data. Before this data unit can be issued to an Ethernet network, it must be segmented into units no larger than 1500 bytes. To learn a network's MTU size (and thereby determine whether it needs to segment packets), Transport layer protocols perform a discovery routine upon establishing a connection with the network. Thereafter, the protocols will segment each data unit as necessary until closing the connection.

Reading Analogy

Segmentation is similar to the process of breaking down words into recognizable syllables that a child uses when learning to read. When a child understands the separate syllables, he can combine them into a word—that is, he can reassemble the parts into a whole.

Route

The process of determining the best path is known as routing. More formally, to route means to intelligently direct data based on addressing, patterns of usage, and availability. Because the Network layer handles routing, routers—the devices that connect network segments and direct data—belong in the Network layer

Encapsulate

The process of wrapping one layer's PDU with protocol information so that it can be interpreted by a lower layer. For example, Data Link layer protocols encapsulate Network layer packets in frames.

Data Link layer has two sublayers

The reason for this change was to allow higher-layer protocols (for example, those operating in the Network layer) to interact with Data Link layer protocols without regard for Physical layer specifications.

Extension Identifier

The remaining characters in a physical address, identify the interface. Vendors such as 3Com and Intel assign each NIC a unique extension identifier, based on the NIC's model and manufacture date. By assigning unique extension identifiers, companies ensure that no two NICs share the same physical address. Extension identifiers may also be known as device IDs

LLC (Logical Link Control) sublayer

The upper sublayer of the Data Link layer, provides an interface to the Network layer protocols, manages flow control, and issues requests for transmission for data that have suffered errors.

Connectionless Protocols

Those that do not establish a connection before transmitting and make no effort to ensure that data is delivered free of errors.

In fragmentation

a Network layer protocol (such as IP) subdivides the segments it receives from the Transport layer into smaller packets. If this process sounds familiar, it's because fragmentation accomplishes the same task at the Network layer that segmentation performs at the Transport layer. It ensures that packets issued to the network are no larger than the network's maximum transmission unit size. However, if a Transport layer protocol performs segmentation, fragmentation may not be necessary. For greater network efficiency, segmentation is preferred. Not all Transport layer protocols are designed to accomplish segmentation. If a Transport layer protocol cannot perform segmentation, Network layer protocols will perform fragmentation, if needed.

ISP (Internet service provider)

a business that provides organizations and individuals with access to the Internet and often, other services, such as e-mail and Web hosting.

For example

a computer running on a TCP/IP network might have a Network layer address of 10.34.99.12 and a physical address of 0060973E97F3. In the classroom example, this addressing scheme is like saying that "Ms. Jones" and "United States citizen with Social Security number 123-45-6789" are the same person.

For Example

a series of Ethernet NICs manufactured by the 3Com Corporation begins with the hexadecimal characters "00608C," while a series of Ethernet NICs manufactured by Intel begins with "00AA00." Some manufacturers have several different OUIs. IEEE also uses the term company_id to refer to the OUI. Traditionally, this portion of a physical address is sometimes called the block ID.

Network layer protocols

accept the Transport layer segments and add logical addressing information in a network header. At this point, the data unit becomes a packet. Network layer protocols also determine the path from point A on one network to point B on another network by factoring in: ● Delivery priorities (for example, packets that make up a phone call connected through the Internet might be designated high priority, whereas a mass e-mail message is low priority) ● Network congestion ● Quality of service (for example, some packets may require faster, more reliable delivery) ● Cost of alternative routes

Acknowledgments

are also used in subsequent communications to ensure that data was properly delivered. For every data unit a node sends, its connection-oriented protocol expects an acknowledgment from the recipient. For example, after a client's TCP protocol issued an HTTP request, it would expect to receive an acknowledgment from the Web server proving that the data arrived. If data isn't acknowledged within a given time period, the client's protocol assumes the data was lost and retransmits it.

Ethernet frames

are different from token ring frames, and the two will not interact with each other on a network. In fact, most LANs do not support more than one frame type because devices cannot support more than one frame type per physical interface, or NIC. (NICs can, however, support multiple protocols.) Although you can conceivably transmit both token ring and Ethernet frames on a network, Ethernet interfaces cannot interpret token ring frames, and vice versa. Normally, LANs use either Ethernet or token ring, and almost all contemporary LANs use Ethernet.

Standards

are documented agreements containing technical specifications or other precise criteria that stipulate how a particular product or service should be designed or performed.

Physical addresses are frequently depicted

as hexadecimal numbers separated by colons—for example, 00:60:8C:00:54:99.Whereas the traditional MAC addressing scheme assigns interfaces a 48-bit address, IEEE's newer EUI-64 (Extended Unique Identifier-64) standard calls for a 64-bit physical address. In the EUI-64 standard, the OUI portion is 24 bits in length. A 40-bit extension identifier makes up the rest of the physical address to total 64 bits.

OSI (Open Systems Interconnection) model

divides network communications into seven layers: Physical, Data Link, Network, Transport, Session, Presentation, and Application.

ISOC (Internet Society)

founded in 1992, is a professional membership society that helps to establish technical standards for the Internet. Some current ISOC concerns include the rapid growth of the Internet and keeping it accessible, information security, and the need for stable addressing services and open standards across the Internet.

ISO (International Organization for Standardization)

is a collection of standards organizations representing 162 countries. ISO's goal is to establish international technological standards to facilitate global exchange of information and barrier- free trade. It also applies to the fields of textiles, packaging, distribution of goods, energy production and utilization, shipbuilding, and banking and financial services. In fact, iso is the Greek word for equal. Using this term conveys the organization's dedication to standards.

Physical Address

is a fixed number associated with a device's network interface. It is assigned to each NIC at the factory and stored in the NIC's on-board memory. Because this address is appended by the MAC sublayer of the Data Link layer, it is also known as a MAC address or a Data Link layer address. Sometimes, it's also called a hardware address.

Sequencing

is a method of identifying segments that belong to the same group of subdivided data. Sequencing also indicates where a unit of data begins, as well as the order in which groups of data were issued and, therefore, should be interpreted.

Ethernet

is a networking technology originally developed at Xerox in the early 1970s and improved by Digital Equipment Corporation, Intel, and Xerox. There are four different types of Ethernet frames. The most popular form of Ethernet is characterized by the unique way in which devices share a common transmission channel, described in the IEEE 802.3 standard.

ITU (International Telecommunication Union)

is a specialized United Nations agency that regulates international telecommunications, including radio and TV frequencies, satellite and telephony specifications, networking infrastructure, and tariffs applied to global communications. It also provides developing countries with technical expertise and equipment to advance those nations' technological bases.

Frame

is a structured package for moving data that includes not only the raw data, or "payload," but also the sender's and receiver's network addresses, and error checking and control information. The addresses tell the network where to deliver the frame, whereas the error checking and control information ensure that the frame arrives without any problems.

EIA (Electronic Industries Alliance)

is a trade organization composed of representatives from electronics manufacturing firms across the United States. EIA not only sets standards for its members, but also helps write ANSI standards and lobbies for legislation favorable to the growth of the computer and electronics industries.

Checksum

is a unique character string that allows the receiving node to determine if an arriving data unit exactly matches the data unit sent by the source. Checksums are added to data at the source and verified at the destination. If at the destination a checksum doesn't match what the source predicted, the destination's Transport layer protocols ask the source to retransmit the data.

Error checking

is accomplished by a 4-byte FCS (frame check sequence) field, whose purpose is to ensure that the data at the destination exactly match the data issued from the source.

ANSI (American National Standards Institute)

is an organization composed of more than a thousand representatives from industry and government who together determine standards for the electronics industry and other fields, such as chemical and nuclear engineering, health and safety, and construction. ANSI also represents the United States in setting international standards.

Physical Layer

is the lowest, or first, layer of the OSI model. Protocols at the Physical layer accept frames from the Data Link layer and generate signals as changes in voltage at the NIC. (Signals are made of electrical impulses that, when issued in a certain pattern, represent information.) When the network uses copper as its transmission medium, these signals are also issued over the wire as voltage. In the case of fiber-optic cable, signals are issued as light pulses. When a network uses wireless transmission, the signals are sent from antennas as electromagnetic waves.

Reassembly

is the process of reconstructing the segmented data units.

When the destination node receives the frame,

its Data Link layer services unscramble the FCS via the same CRC algorithm and ensure that the frame's fields match their original form. If this comparison fails, the receiving node assumes that the frame has been damaged in transit and requests that the source node retransmit the data. Note that the receiving node, and not the sending node, is responsible for detecting errors.

IEEE (Institute of Electrical and Electronics Engineers)

or "I-triple-E," is an international society composed of engineering professionals. Its goals are to promote development and education in the electrical engineering and computer science fields.

Data Link Layer

protocols divide data they receive from the Network layer into distinct frames that can then be transmitted by the Physical layer.

Simple connectivity devices

such as hubs and repeaters operate at the Physical layer. NICs operate at both the Physical layer and at the Data Link layer. As you would expect, physical network problems, such as a severed wire or a broken connectivity device, affect the Physical layer. Similarly, if you insert a NIC but fail to seat it deeply enough in the computer's main circuit board, your computer will experience network problems at the Physical layer.

In traditional physical addressing schemes,

the OUI is six characters (or 24 bits) long and the extension identifier is also six characters long. Together, the OUI and extension identifier form a whole physical address. For example, IBM might assign one of its NICs the extension identifier 005499. The combination of the IBM OUI and this extension identifier result in a unique, 12-character, or 48-bit address of 00608C005499.

IETF (Internet Engineering Task Force)

the organization that sets standards for how systems communicate over the Internet—in particular, how protocols operate and interact. On an international level, IETF works with the ITU to help give technical standards approved in the United States international acceptance.