What You Will Learn
After reading this chapter, you should be able to
Understand the IP version 4 (IPv4) addressing protocol, and the similarities to the addresses used in the postal delivery system
Explain why hosts need two addresses, Ethernet and Internet, and how the Address Resolution Protocol (ARP) is used to determine the Ethernet address of a host given the host's IP address
Understand the TCP and UDP protocols, and how they are used to deliver data to a destination host
Explain the TCP/IP layered model by comparing it to the layer model that was developed for the postal delivery system
Differentiate between classful and classless IP addresses
Learn how to subnet and summarize IP networks
Learn how routers forward packets using the longest match operation
Describe the IP version 6 (IPv6) addressing protocol
In Chapter 1, you examined systems for delivering the mail, planning a road trip, and making telephone calls. Chapter 2 introduced the binary, octal, and hexadecimal numbering systems. You need to understand how computers represent information, and how you can move between number systems to represent binary numbers in a more readable form. In this chapter, the concepts from the first two chapters will be combined to understand the schemes that are necessary to create a scalable computer communication system the Internet.
To begin our discussion on computer communication over a network, this section looks at the similarities between mail delivery between houses, and data delivery between computers. The endpoints in mail delivery are houses, and the endpoints between electronic data delivery are computers. Certainly there can be other endpoints in both systems. Letters can be delivered from a house to a business, from a business to a house, between two businesses, and so on. Electronic data delivery can be from a news service to your cell phone or personal data assistant (PDA), from your computer to your friend's pager, from environmental sensors in a building to the heating and cooling control systems for that building, and so on. But to keep the discussion simple, it will suffice to concentrate on mail delivery between houses, and electronic data delivery between computers. The first analogy is that an endpoint in a mail delivery system, a house, is equivalent to the endpoint in a computer communication system, a PC. (See Figure 3-1.)
Figure 3-1 Equivalent Endpoints in the Mail and Data Communication Systems
In the mail delivery system, the function of the post office is to deliver mail to a particular house. In the computer communication system, the function of the Internet is to deliver data to a particular PC. Yet, in both systems, the endpoint is not the ultimate destination. For mail, the ultimate recipient is a person. For data, the ultimate recipient is an application such as an e-mail program, a web browser, an audio or video program, an instant messaging program, or any number of wonderful applications that exist today. (See Figure 3-2.)
Figure 3-2 Final Destinations in the Postal and Electronic Data Delivery Systems
Although the ultimate recipient is a person or a software application, the responsibility of the systems stops when the mail, or data, is delivered to the proper house, or computer. However, as part of the address, you still need the ultimate recipient; either a person or an application, even though this information is not used for delivery to an endpoint. The endpoint uses the name or application to enable delivery to the recipient.
Because the two systems are analogous, it is instructive to revisit the format of an address in the mail delivery system and see if you can use a similar format for electronic data delivery:
Street Number, Street Name
Although there are five distinct pieces of information in the mail address (name, street number, street name, city, and state), you can consider an address to contain only four pieces of information. For endpoint delivery, you can ignore the name field. You are left with
The postal system routers (core, distribution, and access) use the state, city, and street names to deliver the mail from the source access post office to the destination access post office. The street number is not needed until the mail arrives at the access post office that is directly connected to the destination street. So, the address can be broken down into
State, City, Street Name
The state, city, and street name information enables the mail to get close to the destination (a particular street). The street number is used to deliver the mail to the proper house. What is the analogy in the computer world to houses on a street? Recall from Chapter 1 that a group of computers can directly communicate with each other through a switch residing on a local-area network (LAN). So a LAN is the computer equivalent to a street. (See Figure 3-3.)
Figure 3-3 LAN of Computers Is Similar to a Street of Houses
Chapter 1 also mentioned that computers have an address, and the most common technology used for computer communication is Ethernet. The sample Ethernet address that was presented in Chapter 1 was 00-03-47-92-9C-6F.
Before you learn more about Ethernet addresses, take the following quiz to make sure you understand the concepts described so far:
What number base is used to represent the Ethernet address?
How many bytes are in an Ethernet address?
How many bits are in an Ethernet Address?
How many Ethernet addresses are possible?
Answer:@Hexadecimal, because the symbols C and F are not used in the other number bases that we discussed. Computers compute using binary. The hexadecimal representation is for our benefit because it is easier to read and write.
Answer:@Six. One hexadecimal digit contains 4 bits, or 1/2 bytes. Two hexadecimal digits contain 8 bits, or 1 byte. An Ethernet address contains 12 hexadecimal digits or 6 bytes.
Answer:@48 (8 bits per byte).
Answer:@248 or 281 trillion, 474 billion, 976 million, 710 thousand, 6 hundred fifty-six (281,474,976,710,656).
An Ethernet address is not a property of your PC. An Ethernet address is a property of the Ethernet card, or built in Ethernet port in your PC. If you put a new Ethernet card in your PC, the Ethernet address of your PC changes.
By itself, an Ethernet address cannot deliver data between two endpoints on the Internet. The reason is that there is no structure to an Ethernet address. There are many manufacturers of Ethernet cards for computers, and each manufacturer is assigned a block of Ethernet address to use for their particular brand of card.
An analogy would be to have 281,474,976,710,656 postal addresses that are sold in a local postal address store. Each local postal address store is given a block of numbers from the total range of numbers that are possible. A postal address is just a number between 0 and 281,474,976,710,655. When you build a house, you would go to the local postal address store and your house would be assigned one of the numbers that hasn't yet been assigned. Everyone in your city would need to get a number assigned from the local postal address store. Because people will not be going to the store in any order, numbers will be assigned randomly throughout the city. The only way that these numbers can be used to deliver mail is if every post office at every level (core, distribution, and access) maintained a list of every number, and the route to reach that number. Therefore, every post office would need to maintain a list of 281,474,976,710,656 addresses and the route to get there. Obviously, this is not scalable. So in addition to an Ethernet address, you need another address that has a structure analogous to the structure of the postal address. What you need is an Internet addressing protocol.