The Internet is a communication system that accomplishes one of three broad uses: (1) share resources, (2) share files or data, and (3) communication. The primary applications of the Internet are e-mail, file transfer, the World Wide Web, e-commerce, searches, Voice over Internet Protocol, and video.
E-mail is the exchange of notes, letters, memos, and other personal communication by way of e-mail software and service companies. You write a message to a person and send it over the Internet via your e-mail provider, which, in turn, transfers it to the e-mail provider of the receiving person. That person retrieves the message at his or her convenience. E-mail is one of the most common communication methods in use today and the number one application of the Internet.
File transfer refers to the ability to transfer files of data or software from one computer to another. The file may be text, digitized photographic data, a computer program, and so on. A file transfer program (FTP) allows you to access files on remote computers and download them to your computer, where they may be used. Files can also be “attached” to e-mail messages. File transfer is an excellent research tool because it allows you to access massive amounts of data in the form of books, articles, newspapers, brochures, datasheets, and hundreds of other sources. Apple Computer’s iTunes allows the purchase and download of music.
World Wide Web
Whenever most people refer to the Internet, they are really talking about the World Wide Web (WWW), or the Web for short. The Web is a specialized part of the Internet where companies, organizations, the government, or even individuals can post information for others to access and use. To do this, you establish a website, which is a computer that stores the information you wish to dispense. The information can be on your own computer or on the computer of a Web service provider. The information is presented in the form of pages, a logic subunit that may contain text, graphics, animations, sound, and even video.
You access these websites through the Internet by way of a special piece of software known as a browser. The browser allows you to call up the desired website by name or to search the Internet websites containing information of interest to you. The browser is the software that lets you navigate and explore the Web and access and display the information. The two most widely used browsers are Netscape Navigator and Microsoft Internet Explorer.
A key feature of the Web is hypertext. Hypertext is a method that allows different pages or websites to be linked. When Web pages are created, usually with a language called hypertext markup language (HTML), programmers can insert links to other pages on the website, to other parts of the same page, or even to different websites. For example, a page may contain a highlighted word (usually in blue letters), which means that it is a link to some related topic on a different page or different site. Clicking with the mouse on that word automatically takes you to the related information. This ability to link to related or relevant information is one of the most powerful and useful features of the Web.
E-commerce or electronic commerce refers to doing business over the Internet, usually buying and selling goods and services by way of the Web. Companies selling products or services set up websites describing their wares and offering them for sale. Individuals may buy these items by giving a credit card number or making other forms of payment. The product is then shipped by mail or overnight carrier. Online shopping is expected to grow signifi cantly in the future. Two of the larger e-commerce retailers are Amazon, a book reseller, and eBay, a site where some items are sold by
An Internet search allows a person to look for information on any given topic. Several companies offer the use of free search “engines,” which are specialized software that can look for websites related to the desired search topic. The most commonly used search sites are Google, Yahoo, and Microsoft Bing. You simply access the website through your browser and type in one or more keywords that direct the search engine. Within seconds, the search engine returns a listing of all those sites it finds containing the keywords. Most searches return hundreds and even thousands of references such as websites, advertisements, press releases, magazine and newspaper articles, and many other resources. Voice over Internet Protocol (VoIP). VoIP is the technique of replacing standard telephone service with a digital voice version with calls taking place over the Internet.
Voice is digitized, compressed, and transmitted over the Internet in packets to the called location. There it is decompressed and reconverted to analog. VoIP has largely replaced standard telephone services in companies and governments and has made great inroads into homes across the United States and other countries. Its growth continues.
Video over Internet Protocol
Video or TV over the Internet (IPTV) is also becoming more common. The video (and accompanying audio) is digitized, compressed, and sent via the Internet. This is referred to as the top (OTT) video. It is gradually replacing some video transmitted over the air and by cable television systems. Both IPTV and VoIP make use of the standard Internet connections used for data.
How the Internet Works
The Internet is the ultimate data communication network. It uses virtually every conceivable type of data communication equipment and technique. All the concepts you learned earlier apply to the Internet. Just keep in mind that the information is transmitted as serial binary pulses, usually grouped as bytes (8-bit chunks) of data within larger groups called packets. All the different types of communication media are used including twisted-pair cable, coaxial cable, fiber- optic cable, satellites, and other wireless connections. Internet Addresses. Each individual or computer on the Internet must have some kind of identifier or address. An addressing system for the Internet uses a simplified name-address scheme that defines a particular hierarchy. The upper level of the hierarchy is called a top-level domain (TLD). A domain is a specific type of organization using the Internet. Such domains are assigned a part of the Internet address. The most common domains and their address segments are as follows:
Another part of the address is the hostname. The host refers to the particular computer connected to the Internet. A host is a computer, device, or user on the network. A server provides services such as e-mail, Web pages, and DNS. The hostname is often the name of the company, organization, or department sponsoring the computer. For example, IBM’s hostname is IBM.
The first part of the address is the user’s name or some abbreviation, concatenation, or nickname. You might use your name for an e-mail address or some made-up name that you would recognize. The complete address might look like this: <email@example.com>
The user name is separated from the host by the @ symbol. The host, in this case, XYZ, is the e-mail service provider. Note the dot between the hostname and the domain name. This address gets converted to a series of numbers used by computers on the Internet to identify and locate one another.
To locate sites on the Web, you use a special address called a uniform resource locator (URL). A typical URL is. The first part of the URL specifies the communication protocol to be used, in this case, hypertext transfer protocol (HTTP). The www, of course, designates the World Wide Web.
The abs.com part is the domain or the computer on which the website exists. The item after the slash (/) indicates a directory within the website software. Most websites have multiple directories, which are also usually further subdivided into pages.
A PC is connected to the Internet in a variety of ways. The most common method is through a modem that connects to the telephone system. There are traditional dial-up modems such as those described earlier and the newer asynchronous digital subscriber line (ADSL) and cable TV modems. Wireless modems are also available.
A common way of connecting to the Internet is to use a LAN to which your PC may be connected. Most company and organization PCs are almost always connected to a LAN. The LAN has a server that handles the Internet connection, which may be by way of a T1 line to the telephone system or a fiber-optic connection.
Role of the Telephone System
The familiar telephone system is often the first link to the Internet. Because it is so large and so convenient, it is a logical way to make connections to remote computers. A standard telephone line is used with dial-up modems and DSL modems. In many applications, the telephone system is the only connection between computers. Many corporate computers are connected in this way, and you can connect directly to an online service provider in this way. Its primary function is to connect you to a facility known as an Internet service provider (ISP).
Although most connections to the Internet occur with a modem over the telephone lines, more and more individuals are acquiring broadband connections. A broadband connection is a fast Internet connection provided by a telephone company or a cable TV company. The term fast here means a data transmission speed greater than the speed that can be obtained with a telephone modem, typically 53 kbps. A broadband connection can provide speeds up to many megabits per second.
The most widely used broadband connection is a cable TV modem. Cable TV companies often act as Internet service providers. The cable TV system is ideal for supporting fast data transfers. A special cable modem is provided that connects to a PC by way of an Ethernet interface. Data transfer rates vary depending upon the amount of traffi c on the line, but they can reach as high as 50 Mbps.
The second most widely used broadband connection is the digital subscriber line (DSL). It is most often offered by a local telephone company but may be delivered by an independent supplier. It gives a data rate from 1.5 to 50 Mbps. Internet access is also available through cell phones.
internet Service Provider
An ISP is a company set up especially to tap into the network known as the Internet. It can be an independent company or a local telephone company or a cable TV company. The ISP has one or more servers to which are connected dozens, hundreds, or even thousands of modems, DSLs, or cable connections from subscribers. It is usually the ISP that provides the software you use in communication over the Internet. The ISP also usually provides e-mail service.
There are thousands of ISPs across the United States. Most are set up in medium-size to large cities. Some ISPs are wireless services. The ISP is connected to the Internet backbone by way of a fast digital interface such as via a fiber-optic MAN.
The cloud is a general term for the Internet. In system diagrams, the Internet is sometimes represented by a large white puffy cumulonimbus cloud. The cloud is a massive unseen networking infrastructure that is constantly changing and has never been mapped.
Cloud computing is the process of storing and accessing programs or data over the Internet. Instead of storing your programs and data on your computer’s hard drive, you access the software and storage capability of remote servers via an Internet connection.
All you need is a minimal computer with a Web browser and possibly an applications interface program that lets you use the external resources wherever they may be. E-mail is an example of cloud computing you have probably used. The e-mail programs like Google Gmail and Yahoo mail and e-mail storage are in the cloud and not on your computer.
Cloud computing is a new concept that is gradually replacing the older original form of local computing. The one thing that makes cloud computing possible, practical, and acceptable is very high-speed broadband Internet connections.
There are multiple forms of cloud computing. One of them is a software as a service (SaaS). This is the process of subscribing to a software program. It may be a game, a database, or some application software. Microsoft offers such software.
Another is a platform as a service (PaaS). An example is a development platform that offers all the tools you need to build your own cloud application. This may be a custom software application created for your own use but implemented in the cloud.
Then there is Infrastructure as a service (IaaS). The infrastructure is a company that supplies hardware such as servers or virtual computers and storage units. Users rent remote equipment, facilities, and services to handle their needs. Amazon and Google are examples.
The big question is, why would you give up local control of your stored data and software? There are several benefits. The first is lower cost. The access computer can be a smaller, less powerful computer with a slower CPU, less RAM, and a smaller hard drive—all of which costs less. You may not have to buy servers or large storage units. Second, software costs less. Instead of owning the software, you lease it instead. Third, you can access your software and data from any place at any time if you have access to a laptop, tablet, or smartphone and an Internet connection. Other benefits, from a company or business perspective, are lower costs for IT staff support and less need for physical space and utilities devoted to servers.
There are some insecurities associated with cloud computing. For example, is your data safe from loss or hacking? Can you get access to it anytime you want? Is there a backup? Is your data really private? Do you really own the data, given that it is stored remotely? Will ISP outages prevent you from getting to your data or the software you subscribe to?
Companies and individuals must make up their own minds about the pros and cons of cloud computing, depending on their needs and resources. In any case, cloud computing is a growing phenomenon and will expand greatly in the coming years.
Internet Transmission Systems
The Internet is made up of a huge number of individual elements. These elements include the servers, routers, and transmission media that carry the signals. A huge number of components and interconnections are involved. This section shows the overall operation of the Internet, the transmission protocols used, as well as the most widely used transmission media.
Frame Relay (FR) is a packet-switching protocol standardized by the ITU-T. It packages data to be transmitted into FR frames that have the structure shown in Fig. 15-1. The 8-bit flags signal the beginning and ending of a packet. A two-octet (byte) address field contains all the details regarding the exact destination of the packet through the network. The data fi eld is variable and may contain up to 4096 octets. A 2-octet frame check sequence (FCS) is an error detection code that is compared to the FCS calculated from the received data. If any transmission error occurs, the receiving unit asks for retransmission.
FR is protocol-independent in that it can carry the data from any other transmission method such as Ethernet. Ethernet packets are put into the data field and carried unmodified to the destination. Frame Relay uses existing dedicated digital circuits leased by telecommunications companies. Both T1 and T3 lines are widely used. The equipment used in handling FR is frame relay switches that examine the destination address and forward the packets through one switch after another until the destination is reached. The most common use of FR is in LAN-to-LAN connections where the LANs are widely separated, as in two different company locations.
The FR service was usually offered by local telecommunications carriers, but it is no longer a viable service.
Asynchronous Transfer Mode
Asynchronous transfer mode (ATM) is also a packet-switching system for transmitting data. It uses very short 53-byte packets with a 48-byte data payload and a 5-byte header that designates the destination as well as the type of data to be handled. Any kind of data may be transmitted in this way, including voice, video, and computer data. The packet format is shown in Fig. 15-2.
ATM is transmitted by way of switches that pass the packets from one to the other over the network. An ATM network sets up a virtual circuit that gives the appearance of a single continuous path for the data from source to destination. In reality, the packets most likely take several different paths through the network and may even arrive out of order. The ATM system handles these occurrences such that all data arrives in the desired order.
fiber-optic system. Early ATM networks operated at 155 and 622 Mbps. Today, ATM uses 2.5- and 10-Gbps optical networks.
Frame Relay and ATM are gradually being phased out in favor of the newer Optical Transport Network (OTN) and Ethernet.
The Synchronous Optical Network (SONET) was developed to transmit digitized telephone calls in T1 format over the fiber-optic cable at high speeds. Its primary use is to send time-multiplexed voice or data over switched networks. SONET is used between telephone central offices, between central offices and long-distance carrier facilities, and for long-distance transmission. Until recently, most Internet backbones were SONET point-to-point connections or rings. Today the trend is away from SONET. Most new long-haul networks use Optical Transport Network (OTN) and Ethernet.
SONET is by far the most widely used optical data transmission network in the United States. It is an American National Standards Institute (ANSI) standard as well as a subset of the broader international standard known as the Synchronous Digital Hierarchy (SDH). The latter was developed and sanctioned by the International Telecommunications Union (ITU) and then used throughout the rest of
SONET is a physical (PHY) layer standard that defines the method for formatting and transmitting data in sync with a master timing system keyed to an atomic clock. The standard defines different optical-carrier (OC) speeds, from 51.84 Mbps to 39.812 Gbps. The electric signal to be transmitted is called the synchronous transport signal (STS) and can exist at several levels or speeds, although the base speed (STS-1) is 51.84 Mbps. In the SDH system, the signal is also called the synchronous transport mode, or STM. Fig. 15-3 lists the most common channel speeds.
Earlier systems used OC-3 and OC-12 levels, but today most SONET systems have been upgraded to operate at OC-48. The OC-192 and OC-768 levels are gradually being deployed as technology and economics permit.
SONET is a time-division multiplexing (TDM) transmission scheme that sends time-interleaved data in fixed-length frames of 810 bytes. The frame format consists of nine 90-byte rows. The bytes are transmitted consecutively from left to right and from top to bottom. See Fig. 15-4. In each row, 4 bytes is for overhead, and 86 bytes per row is for data payload. The overhead bytes contain framing, control, parity, and pointer information for managing the payload.
Although SONET can operate in a point-to-point link, the most common topology is a ring. Multiple nodes make up the ring, with an add-drop multiplexer (ADM) at each node. One of the key features and benefits of a SONET system is that the ADM permits data from multiple sources to be added to or extracted from the data stream as required. The ring topology also offers data transmission redundancy, which ensures survivability when a cable has been cut.
The basic SONET transmission speed is OC-1 or 51.84 Mbps. This is the rate achieved when one SONET frame is transmitted in 125 μs. If the higher-speed transmission is desired, as is usually the case, multiple frames are transmitted in this time period by a system of byte interleaving.
Most of today’s long-distance telephone service, as well as the Internet backbone, is made up of a hierarchy of large and small SONET rings. Even though SONET was designed to carry synchronous voice traffic in a circuit-switched environment, its primary function is data transmission with no regard to its content. As a result, SONET can carry asynchronous, packet-switched data from asynchronous transfer-mode (ATM), Frame Relay, Ethernet, or Transport Control Protocol/Internet Protocol (TCP/IP) equipment.
Optical Transport Network
SONET has been adequate for long-haul and metro networks over the past several decades, but it has not kept pace with the demands of the Internet and other networking efforts. Massive Internet access has exploded in recent years. Increased viewing of TV and downloading of videos has further burdened the Internet. Additional stress has come from the new, faster, fourth-generation (4G) smartphones and the attendant need for faster backhaul networks. On top of that, the cloud computing movement requires increasingly faster networks.
The solution is fiber-optic networks. Most of the build-out of long-haul and metro fiber networks in the late 1990s and early 2000s used 2.488-Gbit/s SONET (OC-48) technology. Since then, many systems underwent upgrades to 10 Gbits/s (9.95328-Gbit/s SONET or OC-192) and a precious few to SONET’s peak rate of near 40 Gbits/s (39.81312-Gbit/s SONET or OC-768).
These synchronous systems connect to asynchronous Ethernet networks, creating the need for interface solutions such as Ethernet-over-SONET protocols. With faster Ethernet and Internet Protocol (IP) systems growing in number and the emergence of all-IP 4G wireless networks, SONET has struggled to keep pace.
Add to that the need for speeds beyond 40 Gbits/s, and suddenly a faster IP-based system is no longer a luxury, but rather a necessity. That new system designated to deal with this problem is called the Optical Transport Network (OTN). It is already in place in some locations and starting to expand.
OTN is called the “digital wrapper” or “optical channel wrapper.” It is defined by the ITU-T standards G.872 and G.709. OTN can carry IP traffic with ease, much like SONET and SDH (Synchronous Digital Hierarchy). OTN features line rates of 2.66 Gbits/s (OTU1) to carry SONET OC-48, 10.7/11.09 Gbits/s (OTU2/2e) to carry 10- Gigabit Ethernet or SONET OC-192, and 43.01 Gbits/s (OTU3) to carry SONET OC-768 or 40-Gigabit Ethernet. A-line rate of 112 Gbits/s (OTU4) is also defined to carry 100- Gigabit Ethernet. Overall, OTN is gradually replacing SONET/SDH in most long-haul and metro networks. It is expected that OTN will form most of the Internet backbone, while 40- and 100-Gbit/s optical versions of Ethernet will serve the LANs, data centers, and some metro areas.
OTN is a packet-based technology and was optimized to carry Ethernet as well as SONET. It includes support for the operation, administration, maintenance, and provisioning (OAM&P) feature of SONET that allows the network carrier to monitor and control the network by programming. In addition, unlike SONET, OTN was designed to operate over one or more wavelengths of light on a single fiber. This wavelength division multiplexing (WDM) is described later.
Fig. 15-5 shows the OTN frame format. There are four rows of 4080 bytes (octets) transmitted one after another left to right and top to bottom. The frame is divided into an overhead, payload, and forward error correction (FEC) fields. The Ethernet, SONET, and other protocol frames are mapped into the payload section. Overhead bytes are added to designate addresses, data types, number of data bytes, OAM&P data, and related information.
Finally, an FEC is added. OTN uses an advanced form of the Reed-Solomon FEC designated RS (255, 239). For every 239 bytes of data, 16 additional bytes of parity correction bits are added. This FEC can correct up to 8-bit errors or detect 16-bit errors per row of data. The key benefit of this FEC is that it adds up to 6.2 dB of coding gain, allowing a BER of 10−12 to be achieved. In addition, with this coding gain, OTN can be transmitted over longer spans of fiber without regeneration of the signal. Whereas SONET was limited to a reach of 200 km, OTN has a reach up to 2000 km.*
The router is the single most important piece of equipment on the Internet. You have probably heard of the world’s largest router manufacturer, Cisco. Routers interconnect the various segments of the WAN backbones. Routers are also used to connect MANs to WANs, LANs to other LANs, and LANs to MANs. The routers connect to one another and to the various servers to form a large mesh network connected usually by fiber-optic cable.
A router is an Intelligent computer-like device that looks at all packets transmitted to it and examines their Internet protocol (IP) destination addresses. It then determines the best next path for the data to take to its destination. The router stores information about the other routers and networks to which it is connected and about any nearby networks. This information is stored in a routing table that is compared to the destination address on all incoming packets. Routing algorithms determine the best (closest, fastest) connection and then retransmit the packet. In any given Internet transmission, most packets pass through several routers before they reach their final target computer.
Physically, a router can have several configurations. Any high-speed device can be used that has serial inputs and outputs and a way to examine packet addresses and make decisions regarding routing and switching. For example, a computer with a fast processor, adequate memory, serial I/O, and the appropriate software can serve as a router. Today, most routers are specialized pieces of equipment optimized for the routing function.
The basic configuration of a modern router is shown in Fig. 15-6. It consists of a group of line cards that plug into connectors on a printed-circuit board back plane. The backplane contains the copper interconnecting lines to allow the line cards to transmit and receive data from one another. Transfer speeds are typically many gigabits per second. The line cards communicate with one another through a switch fabric. The switch fabric is a large digital equivalent of a cross bar switch. A cross bar switch is a matrix of switches in a row-column format that allows any one of multiple input lines to be connected to any one of multiple output lines. Superfast electronic switches connect the line cards. See Fig. 15-6.
Each line card has a serial input port and a serial output port. These are usually optical fi berlines, but they could also be RJ-45 Ethernet ports. At the received input, the optical used. Any input may be routed to any output in this way. The processor then passes the packet to the switch fabric where it is sent to an output port. The packet is serialized by a SERDES circuit and then sent to the output port by using SONET, ATM, Ethernet, or OTN.
data is converted to serial electric signals that are then deserialized by a circuit called a serializer/deserializer (SERDES) that performs both serial-to-parallel and parallel-to-serial data conversion. The parallel data is then stored in random-access memory (RAM). This data is then examined by a special processor to determine its destination address. The processor uses a routing lookup table and algorithms that decide which output line is to be used. Any input may be routed to any output in this way. The processor then passes the packet to the switch fabric where it is sent to an output port. The packet is serialized by a SERDES circuit and then sent to the output port by using SONET, ATM, Ethernet, or OTN.
The Internet Backbone
The Internet backbone is a collection of companies that install, service, and maintain large nationwide and even worldwide networks of high-speed fiber-optic cable. The companies own the equipment and operate it to provide universal access to the Internet. In many ways, these companies are the Internet, for they provide the basic communication medium, equipment, and software that permit any computer to access any other.
Although each of the backbone providers has its own nationwide network, the providers are usually connected to one another to provide many different paths from one computer to another. There are more than 50 of these interconnection points in the United States. Known as network access points (NAPs), these facilities provide the links between backbones.
Fig. 15-7 shows a diagram of the main components of the Internet. On the left are three ways in which a PC is connected for Internet access, to the telephone local office via a conventional or DSL modem or by way of a cable modem and cable TV company. The telephone company central office and cable TV company connect to an Internet service provider by way of a local fiber-optic MAN. (Note: The telephone company or cable TV company may also be the ISP.) The ISP contains multiple servers that handle the traffic.
Each connects to the Internet by way of a router that attaches to the Internet backbone at one of the many network access points. The primary equipment at network access points is routers that determine the destination of packets. The “cloud” in Fig. 15-7 represents the Internet backbone. This is one of several large fiber-optic networks, either point-to-point or a ring, that carry the Internet data and interconnect with one another to exchange data.
On the right in Fig. 15-7 are additional connections. Here a company LAN accesses the backbone by way of a regional MAN. A company server contains a website that is regularly accessed by others. Also shown is a Web hosting company that stores the websites and Web pages of others.
The Packet-Switching Transmission System
The Internet is a packet-switching system. Data to be sent is divided up into short chunks called packets and transmitted one at a time. The term datagram is also used to describe a packet. Packets are typically less than 1500 octets long. (Note: In data communication, a byte is more often called an octet, an 8-bit word.) Each packet finds its own way through the complex maze that is the Internet by using addressing information stored in the packet. Not all packets take the same path through the system, and in fact the packets may arrive at the receiving end out of the order in
which they were sent.
Fig. 15-8 shows a simplified version of the packet-switching concept. The packet switching network consists of multiple exchanges with high-speed packet switches that connect to multiple inputs and outputs. The multiple interexchange links produce many possible paths through the network.
Several popular packet-switching systems are in use in Internet backbones. These include Frame Relay, asynchronous transfer mode (ATM), and OTN systems. ATM systems are the fastest and most widely used. They break down all data into 53-byte packets and transmit them over fiber-optic networks at speeds of up to 100 Gbps. ATM is also used to transmit voice and video. In some instances, the packets are packaged into long continuous synchronous data streams and transmitted by way of the SONET. Such transmissions are called packet over SONET (PoS). Newer parts of the backbone use OTN. Most modern backbone networks operate at 40 Gbps or 100 Gbps.
Packet switching requires a set of software protocols that make sure that the data is properly partitioned, transmitted, received, and reassembled. On the Internet, these protocols are called TCP/IP. TCP means Transmission Control Protocol, and IP means Internet Protocol. Both protocols are used to send and receive data over the Internet.
TCP/IP was invented in the 1960s when the Advanced Research Project Agency (ARPA) of the Department of Defense (DoD) tested and established the first packet network. It was first implemented on large computers running the UNIX operating system. Today, TCP/IP is by far the most widely implemented data communication protocol, and it is at the heart of the Internet.
TCP/IP is a layered protocol similar to the OSI seven-layer model discussed in Chap. 11. TCP/IP does not implement all seven layers, although the effect is the same. Some operations of the individual seven layers are combined to form four layers, as shown in Fig. 15-9. The upper, or applications, layer works with other protocols that implement the desired application. The most widely used are the file transfer protocol (FTP), which facilitates the transmission of long files of data; the simple mail transfer protocol (SMTP), which implements e-mail; and the hypertext transfer protocol (HTTP), which provides access to the World Wide Web.
Transmission Control Protocol
The host-to-host layer is really TCP. It divides the data into packets to be sent. To each packet is attached a comprehensive header containing source and destination addresses, a sequence number, an acknowledgment number, a checksum for error detection and correction, and other information. The TCP header is shown in Fig. 15-10. It consists of 20 bytes or octets, minimum, depending upon what options are added. At the receiving end, TCP reassembles the packets in the
proper order and sends them to the application. It also asks for retransmission if a packet is received in error.
TCP is used only to prepare the packets for transmission and reassemble the packets when received. It does not implement the actual packet transmission over the Internet. As you will see, that is the job of IP. TCP’s job is to offer some assurance that the packets arrive in a reliable and high-quality form by providing the error detecting and correcting checksums, sequence numbers, and other features.
The TCP header is a single serial bit stream in which each of the fields is transmitted one after another. This header is difficult to draw and paint as a long single line, so the presentation format in Fig. 15-10 is used. This representation is widely used to illustrate other protocols and headers. In reality, data is transmitted serially from left to right and from top to bottom, row after row, with 4 octets per row.
The TCP packet is then sent to the IP layer, where the IP protocol is used. The IP layer ensures that the packet gets to its destination over the Internet. In Fig. 15-11, an IP header is attached before the resulting new packet is transmitted.
Note in Fig. 15-11 that the destination address has 32 bits. Every computer has one of these addresses assigned to it. The addresses are usually expressed in what is called the dotted-decimal form. The 32-bit address is divided into four 8-bit segments. Each segment can represent 256 numbers from 0 through 255. A typical address would look like this: 126.96.36.199. Dots separate the four decimal segments. These addresses are assigned by an organization known as the Internet Assigned Numbers Authority (IANA). The addresses are subdivided into classes (A through E) to represent different types and sizes of networks, organizations, and individual users.
The most widely used IP protocol is called IPv4, or Internet Protocol version 4. Its primary limitation is that its destination address size is only 32 bits; i.e., it is limited to 232, or 4,294,967,296, locations or users. Although more than 4 billion addresses would seem to be enough, as the Internet has grown, this has become a severe limitation. The newer version is IPv6, which has a 128-bit address field that should provide more than enough destinations (2128) for massive future expansion. Also, IPv6 provides for authentication and multicasting. Multicasting is the ability of IP to move fast audio and video data over the Internet from a single source to multiple destinations. IPv6 is implemented on many computers, but not all. It is slowly being phased in as software and hardware are updated in new systems.
The IPv6 packet with a header is illustrated in Fig. 15-12. The header is a simpler version of the IPv4 packet; only the basic information is included, to make forwarding more efficient. A header extension is used if additional instructions are needed.
Key changes to the header are that there is no header checksum and the time to live field is called the hop limit. This is the maximum allowed number of hops between routers that a packet can make before it is discarded.
Note the 128-bit source and destination addresses. Each address is 16 bytes long and can provide up to 3.4x 1038 addresses. Instead of the dotted-decimal address representation of IPv4, the IPv6 addresses are represented by a 32-digit hexadecimal address. There are eight groups of four hex digits with colons between. An example is 1974:0000:A56E:FFFF:0000:239C:470D:3C09.
As for payload, the payload length field in the header is 16 bits, indicating that the data payload may be up to 65,535 bytes. This is the same as in the IPv4 packet.
Finally, the network access layer that contains the physical layer connection transports the data. As explained earlier, there are several different ways that TCP/IP information is transmitted. For example, the packets may get packaged into Ethernet packets in a LAN and then delivered to a MAN for ultimate connection to the packet-switched WAN backbone. Or the TCP/IP packets may travel via a cable TV modem to an ISP that connects to a MAN via a fiber-optic line and then to the WAN. The WAN may carry the data via ATM, SONET, or OTN through the backbone.
Several protocols or sets of rules are used to communicate over a computer network. Each protocol must accomplish several tasks, such as encapsulation, fragmentation and reassembly, connection control, ordered delivery, flow control, error control, addressing, multiplexing, and transmission services. Perhaps the most common protocol used to accomplish these tasks is the Transmission Control Protocol/Internet Protocol or TCP/IP. TCP/IP is the communications protocol that hosts use to communicate over the Internet, and it establishes a virtual connection between a destination and source host. TCP/IP uses two protocols to accomplish this task, TCP, and IP.
TCP enables two hosts to establish a connection and exchange data. TCP will guarantee both the delivery of data and that the packets will be delivered in the same order in which they were sent. Remember, packets are sent through a network according to the best path. This “best path” choice does not guarantee that all packets will take the same path, nor will they arrive in the same order they were sent. TCP has the job of ensuring that all the packets are received and put back into the correct order before they are passed up the protocol stack.
IP determines the format of the packets. The IP packet format is not discussed in detail, but there are a total of 20 octets in IP version 4 packets. These bits are used to select the type of service, length of the datagram, identification number, flags, time to live, next-higher protocol, header checksum, and various addresses. The packet IP provides a function similar to the address on a postal letter. You write the address on the letter and put it in the mailbox. You and the receiver know where it is sent from and to whom it is being sent, but the path is determined by someone else. That someone else is the routers in the network between the destination and the source. It is TCP/IP that establishes the connection between the destination and the source. TCP steps in and cuts the letter up into smaller pieces or packets and then sends them, ensuring all the packets are received and put back into the proper order.
The User Datagram Protocol (UDP) is another protocol used at the transport level. UDP provides a connectionless service for applications. UDP provides few error recovery services, unlike TCP. However, like TCP, UDP uses IP to route its packets throughout the Internet. UDP is used when the arrival of a message is not absolutely critical. You may recall that from time to time you receive letters for the Current Resident in your mailbox. The senders of this “junk mail” are not concerned that everyone receives the package they send. UDP is similar to the Current Resident mail and is often used to send broadcast messages over a network. A broadcast message is a message that is sent periodically to all hosts on the network in order to locate users and collect other data on the network. UDP messages are also used to request responses from nodes or to disseminate information. Another application of UDP is in the use of real-time applications. With real-time applications, retransmitting and waiting for the arrival of packets is not possible, so TCP is not used for these applications. When real-time data (voice or video) is routed, a connectionless UDP protocol is used. If packets get dropped or fail to arrive, the overall message is usually not corrupted beyond recognition.
Internet and Addressing
You have seen the components that make up a network and how information travels across a network, but how does a packet find its intended destination? The Internet is organized in a hierarchical structure. The entire network is often referred to as the Internet or the World Wide Web. The Internet is subdivided into several smaller networks that are all interconnected by routers.
The Internet connects several separate segments or networks together by using routers. Routers need some way to identify the destination network to which a packet is bound. Routers accomplish this by using the network IP address. All devices on that network share the same network address but have unique host addresses. Packets get routed from network to network until they arrive at the network that contains the host to whom the packet has been sent.
A good example of a hierarchal structure is how large military units are organized. The largest structure is the brigade. The brigade is separated into two regiments with three battalions per regiment. Each battalion has five companies. Each company then has platoons, and each platoon has squads. An individual soldier or sailor is in a squad, platoon, company, battalion, and so on. If you want to contact all the soldiers or sailors in a particular company, you can send a message to just that company. The same is true for a battalion or platoon. In a computer network, the ability to send messages to an individual host on a particular network is also important. Each network is then connected to the entire Internet or the so-called Internet cloud.
We can break each connection to the cloud into its own network, and each network would be connected to the cloud by using a router. Every computer connected to the router is considered to be on the same network.
This arrangement is similar to a family. The router would represent a single-family, say, the Jones family; and all the segments represent children in the Jones family. We can easily identify who is in the Jones family by looking at the last name. A router can recognize who is in its network by using a set of numbers called an IP address.
When a computer receives a packet from the router, the computer first checks the destination MAC address of the packet at the data link layer. If it matches, it’s then passed on to the network layer. At the network layer, it checks the packet to see if the destination IP address matches the computer’s IP address. From there, the packet is processed as required by the upper layers. On the other hand, the computer may be generating a packet to send to the router. Then as the packet travels down the OSI model and reaches the network layer, the destination and source IP address of this packet are added in the IP header.
The format of an IPv4 address is called dotted decimal, and it consists of four numbers from 0 to 255 separated by periods or dots. Each number between the periods is considered an octet because it represents 8 binary bits.
The dotted decimal format is convenient for people to use, but in reality, the router will convert this number to binary, and it sees the above-dotted decimal number as a continuous string of 32 bits. The example below shows an IP address in decimal notation. This IP address (188.8.131.52) is then converted to binary, which is what the computer understands. It’s easier for us to remember four different numbers than thirty-two 0s or 1s.
35 .75 .123 .250
The above IP address would look like 00100011010010110111101111111010 to the
computer. The hexadecimal version is 234B7BFA.
To provide flexibility, the early designers of the IP address standard sat down to sort out the range of numbers that were going to be used by all computers. They organized the IP address into five classes, and we normally use three of these classes. When people apply for IP addresses, they are given a certain range within a specific class depending on the size of their network.
In Table 15-1, you can see the five, classes. The first three classes (A, B, and C) are used to identify workstations, routers, switches, and other devices, whereas the last two classes (D and E) are reserved for special use. The IP addresses listed above are not all usable by hosts!
An IPv4 address consists of 32 bits, which means it is 4 bytes long. The first octet (first 8 bits or first byte) of an IP address is enough for us to determine the class to which it belongs. And depending on the class to which the IP address belongs, we can determine which portion of the IP address is the network ID and which is the host ID. For example, if you were told that the first octet of an IP address was 168, then, using the above table, you would notice that it falls within the 128 to 191 range, which makes it a class B IP address.
Earlier we said that companies are assigned different IP ranges within these classes, depending on the size of their networks. For instance, if a company required 1000 IP addresses, it would probably be assigned a range that falls within a class B network rather than a class A or C.
To get the information to the correct host, the IP address is divided into two parts, the network ID and the host ID. These two parts contain two pieces of valuable information: It tells us which network the device is part of (network ID).
It identifi es that unique device within the network (host ID).
- It tells us which network the device is part of (network ID).
- It identifies that unique device within the network (host ID).
Fig. 15-13 gives you an example to help you understand the concept.
Routers will look at the first number or octet to determine the class of the IP address. The class indicates how many bits are used to represent the network ID and how many bits are used to represent the host ID. In the above picture, you can see a small network. We have assigned a class C IP range for this network. Remember that class C IP addresses are for small networks. Looking now at host B, you will see that its IP address is 184.108.40.206. The network ID is 202.178, and the host ID is 0.3.
Table 15-2 contains the range of numbers used to determine the class of the network and the number of bits available to assign to a network and the hosts on that network.
For example, 220.127.116.11 is a class B address. The 140 falls within the 128 to 191 range, which makes it a class B IP address. So, by default, the network part of the address (also known as the network address) is defined by the first 2 octets (140.179.x.x), and the node part is defined by the last 2 octets (x.x.220.200).
Now we can see how the Class determines, by default, which part of the IP address belongs to the network (N) and which part belongs to the host (h).
- Class A—0NNNNNNN.hhhhhhhh.hhhhhhhh.hhhhhhhh
- Class B—10NNNNNN.NNNNNNNN.hhhhhhhh.hhhhhhhh
- Class C—110NNNNN.NNNNNNNN.NNNNNNNN.hhhhhhhh
Take a class A IP address as an example to understand exactly what is happening. Any class A network has a total of 7 bits for the network ID (bit 8 is always set to 0) and 24 bits for the host ID. Now, all we need to do is to calculate how much is 7 bits: 27 =128 networks, while 224 = 16,777,216 hosts in each network. Of the 16,777,216 hosts in each network, 2 cannot be used. One is the network address, and the other is the network broadcast address (see Table 15-3). Therefore when we calculate the “valid” hosts in a network, we always subtract 2.
I asked how many valid hosts you can have on a class A network, you should answer 16,777,214, not 16,777,216. The same story applies to the other two classes we use, class B and class C. The only difference is that the number of networks and hosts changes because the bits assigned to them are different for each class. So if asked how many valid hosts you can have on a class B network, you should answer 65,534, not 65,536. And if asked how many valid hosts you can have on a class C network, you should answer 254, not 256.
Now you’ve learned that even though we have three classes of IP addresses that we can use, some IP addresses have been reserved for special use. This doesn’t mean you can’t assign them to a workstation; but if you did, it would create serious problems within your network. For this reason, it’s best to avoid using these IP addresses. Table 15-3 shows the IP addresses that you should avoid using.
It is imperative that every network, regardless of class and size, have a network address (first IP address, for example, 18.104.22.168 for class C network) and a broadcast address (last IP address, for example, 22.214.171.124 for class C network), as mentioned in Table 15-3 and the diagrams above, which cannot be used. So when you calculate available IP addresses in a network, always remember to subtract 2 from the number of IP addresses within that network.
Reserved Host ID Numbers
When you design a network, you will be given a network ID from a controlling authority. The network ID portion of your Internet address cannot change, but the host ID a portion of the IP address is the bits you own, and you can assign any IP within the host ID portion to the computers on your network, other than the all-0s or all-1s host ID mentioned above. Typically designers will assign the first assignable IP to the first host, but there is no restriction and any number can be assigned.
Just as the name Jones identifies the family members, the network ID identifies your network, but how does the router figure out that the network ID is a match? Information travels in packets, and each packet has a header. The header will contain the IP address of the computer to which the packet is being sent. The router will use a special sequence of bits called the network mask to determine if the packet is being sent to its network. The network mask has all 1s in the network ID and all 0s in the host ID. This mask is then logically ANDed to the packet, and the router will see if the destination host is on its network. In our 126.96.36.199 network, the network mask would be 255.0.0.0. If the network were a class B, the network mask would be 255.255.0.0, and for a class C, the network mask would be 255.255.255.0.
Table 15-4 shows our three network classes with their respective network masks. An IP address consists of two parts, the network ID and the host ID. We can see this once again shown below, where the IP address is analyzed in binary, because this is the way you should work when dealing with network masks:
This class C network uses 21 bits for the network ID (remember, the first 3 bits in the first octet is set) and 8 bits for the host ID. The network mask is what splits the network ID and host ID.
We are looking at an IP address with its network mask for the first time. What we have done is to take the decimal network mask and convert it to binary, along with the IP address. It is essential to work in binary because it makes things clearer and we can avoid making mistakes. The 1s in the network mask are ANDed with the IP address, and the result defines the network ID. If we change any bit within the network ID of the IP address, then we immediately move to a different network. So in this example, we have a 24-bit network mask (twenty-four 1s, counting from left to right).
As the network grows, it becomes increasingly difficult to efficiently route all the traffic, since the router needs to keep track of all the hosts on its network. Let’s say that we have a simple MAN of two routers. Whenever a packet is sent from one host to another on our network, the router will route the packet to the proper host. When the packet is sent to a host connected to another router, the network mask will be used to determine that the packet is not for our network, and the router will send the packet over the communication link to the other network. If each router has only a few hosts connected to it, then the number of packets that the router has to route is relatively small. As the network grows, this number of hosts gets quite large. For a class A network, there could be 224, or close to 17 million, hosts. As a means to help the routers more efficiently route
packets and manage the size of their router tables, a technique called subnetting is used.
When we subnet a network, we basically split it into smaller networks. By subnetting the network, we can partition it into as many smaller networks as we need. By default, all types of classes (A, B, and C) have a network mask. However, a network mask other than the default can be used. This is called a subnet mask. The use of an IP address with a subnet mask other than the network default results in the standard host bits (the bits used to identify the host ID) being divided into two parts: a subnet ID and host ID. Take the same IP address as above, and divide it further into a subnet ID and host ID, thus changing the default network mask. Then by using the ANDing technique described earlier, smaller subnets can be created and identified as required.
MAC Address Versus IP Address
Recall the media access control (MAC) address discussed earlier. The MAC address is a unique address assigned to the physical device. The IP address is a logical address used to determine where in the network a host is located. As in the postal example, the state could be considered the network, with the city the subnet, and the individual person as the host in the network. The MAC address is like the Social Security number (SSN) that each person has. The SSN gives no information about where the person is located but does uniquely identify that person. The MAC identifies the manufacturer and has a unique number associated with it. The IP address is used to find out where the MAC is so that packets can be routed to the host.
An Example Transmission
Assume an e-mail application in which a message packaged by the SMTP is sent to the TCP layer for formatting. The TCP is then sent to the IP layer, which further packages the message for transmission. Also, assume that the IP packets are encapsulated into Ethernet LAN packets and sent to a WAN. At the first WAN node, the router strips off the Ethernet packets recovers the IP header, and reveals the destination. The router makes its routing decision and then re-encapsulates the packets into the protocol of the WAN for this application OTN. The OTN packets then travel from router to router, where the same thing occurs. The packets finally reach the final destination, where the router again recovers the IP packets.
At the receiving end, the IP layer verifies that the packet has come to the right place and then strips off the header and passes it to the TCP layer. Here any errors are detected and retransmission is requested if necessary. Then all the packets are reassembled in the correct order, the header is removed, and the final data is sent to the applications layer, where it is sent to the SMTP protocol that delivers the e-mail.
Radio Wave | Reflection | Refraction | Diffraction | Ground, Sky Waves ( Internet Working | SONET | TCP/IP | Internet Transmission Systems )
Mcqs of Electronics | online Test of Electronics | Thevenin’s | Nortons ( Internet Working | SONET | TCP/IP | Internet Transmission Systems )
Antenna | Antenna Operation | Antenna Types | Radio Waves | Dipoles ( Internet Working | SONET | TCP/IP | Internet Transmission Systems )
Smith Chart | Wavelength Scales | SWR Circle | Plotting and Reading ( Internet Working | SONET | TCP/IP | Internet Transmission Systems )
Standing Waves | Matched Lines | Circuit Elements | Stripline | ( Internet Working | SONET | TCP/IP | Internet Transmission Systems )
Active Filters | Crystal Filters | Ceramic Filters | Surface Acoustic Wave ( Internet Working | SONET | TCP/IP | Internet Transmission Systems )
Click Here to Learn More ( Internet Working | SONET | TCP/IP | Internet Transmission Systems )
Click Here to Learn ( Internet Working | SONET | TCP/IP | Internet Transmission Systems )