Reference Series Table of Contents For This Issue

How Computers Work, Part I
August 2001• Vol.5 Issue 3
Page(s) 180-187 in print issue

Inside The Internet
How Information Travels Online
Most of the time, you couldn’t care less about how the Internet works. As long as you’re getting your e-mail and you can find the latest baseball scores on the Web, you’re happy.

It’s those times when your Internet connection is broken in the middle of a three-hour file download or when the Web site you want to visit delivers an error message that you probably exclaim, “How does this stupid Internet work, anyway?”

The Internet is actually an amazing collaboration of thousands of networks around the world, sharing information using a single protocol. In 30-plus years, the Internet has grown from a four-computer network to an estimated 400 million users or more. Yet the protocols developed in the Internet’s early stages continue to be used today.

  Physical Components. You’ll discover that exact data about the Internet, such as the number of users or the exact routing of cables, is impossible to find. That’s because no single entity owns or oversees the Internet, making it tough to track all of the components or count the users.

The components that make up the Internet are owned and shared by thousands of public and private entities. The supercomputers, or NAPs (Network Access Points) that serve as the major hubs of the Internet are owned by several companies, including MCI WorldCom and AT&T. The connections spread out from the NAPs to routers (Internet-dedicated computers that move Internet traffic) and regional networks through high-speed cables, called backbones. The NAPs are connected to each other by backbones. Spreading out from the regional networks are smaller networks, private networks, and ISPs (Internet service providers). Your Internet connection almost certainly will come through an ISP or through a private network at your company, governmental entity, or educational unit.

The components that make up the Internet come in a variety of forms. Some computers are dedicated to routing Internet traffic, while others share their resources between the Internet and private networks. Some cables that serve to connect remote offices for a single company also carry Internet traffic. The private companies that own the routers and Internet components have a working agreement stating they’ll carry each other’s traffic to make the Internet work as smoothly as possible.

  Cabling. Smaller networks and individual ISPs handle far less Internet traffic than the major routers or NAPs, and the cable connecting them is much smaller than the backbone cables. Several sizes of cables connect the various components of the Internet.

Smallest. These cables, normally used by individuals and small businesses to connect to the Internet, include common telephone lines. Speeds ranging from 14.4Kbps (kilobits per second) to 128Kbps are common.

Small. Outside of common telephone lines, individual users have several options for faster Internet connections, including cable television lines, satellite television connections, and digital telephone lines (such as DSL [Digital Subscriber Lines]). Speeds usually range from 128Kbps to 8Mbps (megabits per second), or about 8 million bits per second.

Medium. Companies using ISPs to connect to the Internet tend to use cables dedicated to carrying Internet traffic to make the connection. Some of the cables in use are T1, T2, and frame relay. Speeds usually range from 512Kbps to 20Mbps.

Large. As the ISPs connect to the Internet, they need high-speed, dedicated cables to carry the traffic from all of their customers. While some smaller ISPs use twisted-pair or coaxial cable to make this connection, others use optical fiber. Large companies that connect their networks directly to the Internet also may use these large cables, including T3, T3D, or OC1. Speeds usually range from 8Mbps to 135Mbps.

Largest. The cables used to create the Internet backbone, connecting the NAPs to routers and to each other, are the largest and fastest cables. Optical fiber is used for the largest cables, which include OC3, OC12, OC24, and OC256. Speeds usually range from 125Mbps to 13Gbps (gigabits per second), or about 13 billion bits per second.

  TCP/IP. The key to downloading and sharing data across the Internet in a format that may be read and deciphered by almost any computer, no matter what OS (operating system) it’s running, is TCP/IP, or Transmission Control Protocol/Internet Protocol. TCP is the protocol used to break the data into packets, which are the largest blocks of data that can be sent across the Internet (less than 1,500 characters or 128KB). IP is the protocol used to send the data across the Internet to its final destination. TCP is then used again to reassemble the packets in the correct order. TCP also checks the packets for any errors that may have occurred during transmission.

The TCP and IP protocols work together to transmit data from one Internet destination to another, whether you’re sending an e-mail message to your mom or accessing the latest stock quotes from your Web broker. Other protocols may work inside the parameters of TCP/IP, depending on the type of Internet usage.

When one computer is told to send data across the Internet, TCP is applied to the data. TCP creates each packet, adding a header and a checksum (a calculated value transmitted with data to check for errors in the transmission) to each packet. The header specifies the order in which the packets must be reassembled; the checksum specifies the exact amount of data in the packet. This is used to check for errors.

Next, IP is applied to each packet, labeling each packet with its destination address. You can think of the IP process as placing each packet inside an envelope with the destination address printed on the outside.

The packets are then sent along the Internet network (a group of computers that are connected and can share data because they’re using the same protocol or OS). Because the Internet is the ultimate network—often called a network of networks—the packets may encounter several different types of computers running several different OSes. TCP/IP is the common thread, though. Just as computers in a corporate network can communicate because they’re running the same OS, all computers on the Internet can communicate because they can all use TCP/IP. (For more on networking, see Get Connected With Networks" in this issue.)

As the packets travel along the Internet, they encounter routers. The routers read the IP address on each packet. If the packet is destined for the regional network or local network served by the router, it directs it to the final destination. Otherwise the router must send the packet along to the router closest to the packet’s final destination.

The router uses a router table to determine where to send the packets. A router table is a database stored in the router with information about which routers should handle certain addresses. Some routers, called static routers, always send the packets in the same direction. Other routers, called dynamic routers, can choose from several different routes, bypassing busy routers to move the packets more quickly. You can think of the static router as a commuter who always drives the same route to work, regardless of the rush-hour consequences. A dynamic router would be similar to a commuter who listens to radio reports on traffic congestion before choosing the best route to work.

During the busiest times of Internet usage, you probably notice some slowing in Web page loading or file downloading. One explanation is the routers are receiving requests to move packets faster than they can handle them, causing some packets to be lost. When a packet is lost, it must be re-sent, causing additional traffic and additional delays.

Once the packets arrive at their final destination, TCP is used to check each of them for errors, using the checksum. If any errors are detected, a request is made to have the packet re-sent. After all of the packets arrive, TCP is used to arrange them in the correct order, making them readable. (For more information, see the “How Information Travels Through The Web” infographic.)

  The Web. Without TCP/IP, transferring data between computers on the Internet would be nearly impossible, no matter what portion of the Internet you were using, including the Web.

The World Wide Web (often shortened to WWW or Web) is probably the portion of the Internet with which you’re most familiar; the Web has become the commercial arm of the Internet. Whether you want to purchase a book, research a stock, or find an archived Smart Computing story, the Web is probably where you’ll turn. The Web’s ability to deliver a mixture of graphics, photos, audio clips, video clips, and text make it a powerful tool.

To use the Web, you must first make an Internet connection and open your Web browser (the software that lets you load and view documents from the Web). After you type the URL (uniform resource locator) of the Web page and press ENTER, a complex process takes place behind the scenes to load the Web page. The address is sent across the Internet using a combination of TCP/IP and HTTP (Hypertext Transfer Protocol), the protocol used by Web browsers and servers to communicate.

Each portion of the Web address is important to loading the proper page. The http: at the beginning of the address signifies the use of HTTP. The www signifies use of Web technology. The or portion signifies the exact Web server that contains the desired Web page. The portion of the address after the .com or .org specifies a particular page or directory on the server. For details on how hypertext works, see Connecting The Dots in this issue.

Web routers move the address along the Internet, steering it toward the correct server. Once the server receives the address, it sends the Web page and any accompanying graphics and text back across the Internet to the computer that made the request. The Web browser accepts all of the elements and assembles them in the correct order to be displayed.

Once the Web server completes your request, it moves on to another request; it doesn’t keep a connection open for any additional requests from you. If you later make another request from the same Web server to load another page, the entire process begins again. However, subsequent pages may load more quickly because some of the elements of the page are already stored in the Web browser’s cache memory on your computer.

  Internet Information. The amount of data on the Internet is overwhelming to consider. But who creates this data? Here are some of the main groups that provide Internet data.

This backbone map from UUNET (an MCI WorldCom company) provides a look at the cabling structure of a portion of the Internet in North America.
Commercial. Companies see the Internet, and especially the Web, as a large money-making market where they can sell products. Some Web-based companies, such as, already have established themselves. Such Web stores—and the Internet advertising that accompanies them—should continue to provide a large chunk of Internet data.

Nearly every major company in the world has created a Web site, but some of them simply provide product information as opposed to selling products. Customers can also find technical support for products on the Internet. Visitors seeking employment with a company can often find job openings and apply at the company’s Web site.

Some companies have created intranets and extranets, which use TCP/IP protocols and Internet technology to create networks within the company and provide information to remote and in-house employees.

Government. Federal, state, and local governments provide much of the Internet data, giving individuals and companies access to forms and information. For instance, on the Internal Revenue Service Web site, taxpayers can download tax forms or find income tax information.

Education. Educational institutions were some of the first Internet users, and they continue to generate plenty of Internet data. Researchers working across the world can exchange data through the Internet. Professors can teach classes through the Internet without seeing the students face-to-face. With the large influx of commercial applications on the Internet, however, they are finding unacceptable slowdowns in Internet traffic, prompting the development of Internet2.

Individuals. One of the most appealing aspects of the Internet is its ability to allow anyone to publish anything. Individuals now have an unmatched ability to reach thousands of people with a published Web page. The decision to reach visitors with a theory on a UFO conspiracy or with the recipe for grandma’s chocolate chip cookies is, of course, up to the individual.

  Navigating The Internet. So how do you get to that recipe? Connecting to the Internet is easy using modern equipment. There are a variety of ways to access the Internet, and you don’t even need to own a computer.

Modem. The most popular way for individuals to gain access to the Internet is through a computer modem connected to a telephone line. (A modem is a piece of hardware that translates computer data into signals that can be transmitted across telephone lines.) The modem dials your ISP’s modem, making the connection. The ISP’s computer serves as your gateway to the Internet, transferring your data requests to the proper server on the Internet.

To read Web data or to send and receive e-mail, you’ll need Web browsing software.

LAN. Many companies allow employees to connect to the Internet through the company’s LAN (local-area network). Larger companies often have an Internet server available that manages all active Internet connections along the network. Such companies usually use a dedicated cable to keep the Internet connection available at all times. As long as you’re connected to the network and you have the right software, you can access the Internet through the LAN.

Other access devices. If you don’t own a computer but still want Internet access, you can use a product such as Microsoft’s WebTV, which connects to the Internet through a set-top box and telephone line. It turns your television into a monitor that displays Internet data.

  The Future. Considering the evolution of the Internet and the increase in traffic during the 1990s, the ability of the Internet to thrive using many of its original protocols is remarkable. The developers of the Internet may not have foreseen all of today’s uses for the technology, but they did plan for most of them.

How long the Internet will continue to operate in its current form is impossible to predict. As Internet2 develops and as new protocols are created, they’ll probably begin to filter down to the original Internet, expanding its ability to deliver data more efficiently and more quickly.

Until then, Internet users will have to suffer through the occasional irregularity or quirk. Now that you know the hoops the millions of packets must travel through before reaching your computer, you probably can forgive the occasional error message.  

by Kyle Schurman

View the graphics that accompany this article.
How The Internet Works
How Information Travels Through The Web
(NOTE: These pages are PDF (Portable Document Format) files. You will need Adobe Acrobat Reader to view these pages. Download Adobe Acrobat Reader)

History Of The Internet

For the majority of current Internet users, the Internet didn’t “arrive” until the mid-1990s when the Web made it accessible to the masses. But well before you could purchase a computer at the Gateway Web site or visit the Tide Web site for the latest laundry tips, the Internet was a key component of national defense in the United States.

The Internet was developed in the mid-1960s because the ARPA (Advanced Research Projects Agency) of the DOD (Department of Defense) needed a means of exchanging data between its remote offices. The DOD was also concerned about the ability of a single missile attack to destroy a main computer center. By sharing data between computers at several locations and by having several lines of communication available to share the data, a single attack couldn’t knock out the entire network.

ARPA originally developed the concept of using one computer, or router, to move the data using a universal protocol, while still allowing the other computers on the network to run their unique operating systems. While the idea of TCP/IP (Transmission Control Protocol/Internet Protocol) came into play in the early 1970s, the official TCP/IP protocols weren’t accepted until later that decade. In 1991, the World Wide Web was launched, using TCP/IP to exchange data.

In the original idea of the Internet, called ARPAnet, the only type of planned transmission was for data. Video and audio over the Internet were almost incomprehensible at that time. Security was not a major issue because the only plans were for sharing research among peers.

Over the years, though, the Internet has seen explosive growth. In 1969, three nodes in California and one node in Utah were connected. Three years later, 29 nodes were connected. By 1977, about 100 computers were connected, and by 1983, nearly 1,000 were connected. The Internet topped 100,000 host computers in 1990. Now there are more than 400 million users worldwide.

Over the years, management of the Internet shifted from ARPA to the Defense Communications Agency and finally to the NSF (National Science Foundation). In 1994, the NSF and the federal government decided to remove themselves from the management of the Internet, and it developed into the collection of private networks connected by routers that it is today.

Want more information about a topic you found of interest while reading this article? Type a word or phrase that identifies the topic and click "Search" to find relevant articles from within our editorial database.

© Copyright by Sandhills Publishing Company 2001. All rights reserved.