Computer Science W5 Assignment

profileCyberSter
The_Internet_Book_Everything_You_Need_to_Know_abou..._----_Chapter_9_Packet_Switching.pdf

9

Packet Switching

9.1 Introduction

This chapter begins an exploration of the basic communication technology that the Internet uses. It describes the fundamental mechanism all computer networks use to transfer data, and explains why the scheme works well. Succeeding chapters show how the Internet uses the mechanism. Understanding how networks function is important because knowing about technology allows one to appreciate mechanisms, understand possibilities, and distinguish between apparent magic and advanced technology.

9.2 Sharing To Reduce Cost

Chapter 7 claims that using a dedicated connection between each pair of computers is too expensive. To understand the expense, look at some examples. If four devices need a wired connection between each possible pair, a total of only six connections are needed. For seven devices, the total is twenty-one, and for twenty devices, one hundred ninety wires are required. If an organization has fifty devices, over twelve hundred con- nections are needed!

To avoid the expense and inconvenience of running a dedicated connection between each pair of communicating devices, a computer network arranges for multiple devices to share the underlying transmission facilities. We can summarize:

97

9

Packet Switching

9.1 Introduction

This chapter begins an exploration of the basic communication technology that the Internet uses. It describes the fundamental mechanism all computer networks use to transfer data, and explains why the scheme works well. Succeeding chapters show how the Internet uses the mechanism. Understanding how networks function is important because knowing about technology allows one to appreciate mechanisms, understand possibilities, and distinguish between apparent magic and advanced technology.

9.2 Sharing To Reduce Cost

Chapter 7 claims that using a dedicated connection between each pair of computers is too expensive. To understand the expense, look at some examples. If four devices need a wired connection between each possible pair, a total of only six connections are needed. For seven devices, the total is twenty-one, and for twenty devices, one hundred ninety wires are required. If an organization has fifty devices, over twelve hundred con- nections are needed!

To avoid the expense and inconvenience of running a dedicated connection between each pair of communicating devices, a computer network arranges for multiple devices to share the underlying transmission facilities. We can summarize:

97

K392894.indd 125 03/08/18 12:07 pmComer, Douglas E.. The Internet Book : Everything You Need to Know about Computer Networking and How the Internet Works, CRC Press LLC, 2018. ProQuest Ebook Central, http://ebookcentral.proquest.com/lib/apus/detail.action?docID=5502790. Created from apus on 2025-02-06 02:33:15.

C op

yr ig

ht ©

2 01

8. C

R C

P re

ss L

LC . A

ll rig

ht s

re se

rv ed

.

98 Packet Switching Chap. 9

Because running connections between every pair of communicating devices is prohibitively expensive, networks arrange for multiple com- munications to share a given transmission path.

9.3 Sharing By Taking Turns

How can multiple computers share a transmission path? They take turns. A good analogy comes from early telephone systems. Subscribers who wanted to lower costs could choose a party-line service. Instead of running a separate set of wires to each subscriber’s house, a party-line service meant one set of wires ran down the street and multiple telephones connected to the set of wires. If one of the subscribers on the party line was using the phone, another user on the party line who picked up their phone would hear the conversation. Etiquette dictated that if a subscriber heard someone else talking, they were to hang up immediately and try again later.

Many computer networks use the party-line approach — multiple computers share a transmission system. When a computer has data to send, the computer sends immedi- ately if the network is idle, and must wait for the transfer to complete if the network is in use. The scheme is used for wireless networks, such as Wi-Fi, as well as wired net- works. The point is:

Only one data transfer can occur on a transmission path at a given time. When computers share a network, they take turns sending data.

9.4 Avoiding Long Delays

We have all experienced the downside of taking turns to share a resource: delay. For example, consider an office where employees share a photocopier. Imagine the frustration of needing to copy one page and finding that you must wait for someone who is copying a 900-page document.

If computer networks followed the same approach, long transfers would leave users frustrated. For example, suppose all the subscribers on your street shared a connection that led to your ISP, and suppose all the subscribers took turns using the connection. Imagine how frustrating it would be if you had to wait twenty minutes while a neighbor streamed an HD movie before you could use the connection.

To avoid situations where the data transfer by one device leaves others waiting, networking researchers invented a system that prevents long delays. The idea is straightforward: instead of allowing a given device to use the network for an arbitrarily long time, limit the amount of data that a device can transfer on each turn. The idea, which was invented in the 1960s, is called packet switching, and the unit of data that can be transferred is called a packet. Figure 9.1 illustrates how devices use packet switching.

Sec. 9.4 Avoiding Long Delays 99

The figure shows four devices attached to a network. Device A is sending data to Device C, while Device B sends data to Device D. Each sender divides its outgoing message into packets, and they take turns sending packets. First A sends a packet, then B sends a packet, then A sends a packet, and so on.

A

B

C

D

CDCDC

1st packet from A to C 1st packet from B to D

2nd packet from A to C 2nd packet from B to D 3rd packet from A to C

Shared Network System

Figure 9.1 An illustration of devices taking turns sending packets across a network. Device A communicates with device C, while device B communicates with device D.

9.5 Long Messages And Short Packets

Both LANs and WANs use packet switching, as do both wired and wireless net- works. Furthermore, packet switching is used with all applications. Whether you send a text message, photo, document, video clip, or download a song, the data is always sent in packets. When a user requests a data transfer, software on the sending device divides the item to be sent into packets before sending; on the receiving device, software collects the series of incoming packets, and reconstructs the original item. A user remains unaware that packets are being used.

The maximum size of a packet is set by a network designer. For Ethernet, the maximum size is 1500 bytes of data; for Wi-Fi, 2304 bytes of data can be sent in a sin- gle packet. The details are not important, but remember that when you transfer a large item (e.g., a movie), the item is divided into many packets. If an item is short enough, the entire item can fit into a single packet. Because devices take turns, a single-packet message can be sent without waiting for a long transfer to complete. From a user’s point of view, a short message appears to “sneak in” and use the network while a long transfer continues.

9.6 Each Packet Contains Extra Information

Each packet sent across a network originates at one device and is destined for another. If devices take turns sharing, how can the network hardware tell which device should receive a given packet? The answer is that computer networks operate the same

K392894.indd 126 03/08/18 12:07 pmComer, Douglas E.. The Internet Book : Everything You Need to Know about Computer Networking and How the Internet Works, CRC Press LLC, 2018. ProQuest Ebook Central, http://ebookcentral.proquest.com/lib/apus/detail.action?docID=5502790. Created from apus on 2025-02-06 02:33:15.

C op

yr ig

ht ©

2 01

8. C

R C

P re

ss L

LC . A

ll rig

ht s

re se

rv ed

.

98 Packet Switching Chap. 9

Because running connections between every pair of communicating devices is prohibitively expensive, networks arrange for multiple com- munications to share a given transmission path.

9.3 Sharing By Taking Turns

How can multiple computers share a transmission path? They take turns. A good analogy comes from early telephone systems. Subscribers who wanted to lower costs could choose a party-line service. Instead of running a separate set of wires to each subscriber’s house, a party-line service meant one set of wires ran down the street and multiple telephones connected to the set of wires. If one of the subscribers on the party line was using the phone, another user on the party line who picked up their phone would hear the conversation. Etiquette dictated that if a subscriber heard someone else talking, they were to hang up immediately and try again later.

Many computer networks use the party-line approach — multiple computers share a transmission system. When a computer has data to send, the computer sends immedi- ately if the network is idle, and must wait for the transfer to complete if the network is in use. The scheme is used for wireless networks, such as Wi-Fi, as well as wired net- works. The point is:

Only one data transfer can occur on a transmission path at a given time. When computers share a network, they take turns sending data.

9.4 Avoiding Long Delays

We have all experienced the downside of taking turns to share a resource: delay. For example, consider an office where employees share a photocopier. Imagine the frustration of needing to copy one page and finding that you must wait for someone who is copying a 900-page document.

If computer networks followed the same approach, long transfers would leave users frustrated. For example, suppose all the subscribers on your street shared a connection that led to your ISP, and suppose all the subscribers took turns using the connection. Imagine how frustrating it would be if you had to wait twenty minutes while a neighbor streamed an HD movie before you could use the connection.

To avoid situations where the data transfer by one device leaves others waiting, networking researchers invented a system that prevents long delays. The idea is straightforward: instead of allowing a given device to use the network for an arbitrarily long time, limit the amount of data that a device can transfer on each turn. The idea, which was invented in the 1960s, is called packet switching, and the unit of data that can be transferred is called a packet. Figure 9.1 illustrates how devices use packet switching.

Sec. 9.4 Avoiding Long Delays 99

The figure shows four devices attached to a network. Device A is sending data to Device C, while Device B sends data to Device D. Each sender divides its outgoing message into packets, and they take turns sending packets. First A sends a packet, then B sends a packet, then A sends a packet, and so on.

A

B

C

D

CDCDC

1st packet from A to C 1st packet from B to D

2nd packet from A to C 2nd packet from B to D 3rd packet from A to C

Shared Network System

Figure 9.1 An illustration of devices taking turns sending packets across a network. Device A communicates with device C, while device B communicates with device D.

9.5 Long Messages And Short Packets

Both LANs and WANs use packet switching, as do both wired and wireless net- works. Furthermore, packet switching is used with all applications. Whether you send a text message, photo, document, video clip, or download a song, the data is always sent in packets. When a user requests a data transfer, software on the sending device divides the item to be sent into packets before sending; on the receiving device, software collects the series of incoming packets, and reconstructs the original item. A user remains unaware that packets are being used.

The maximum size of a packet is set by a network designer. For Ethernet, the maximum size is 1500 bytes of data; for Wi-Fi, 2304 bytes of data can be sent in a sin- gle packet. The details are not important, but remember that when you transfer a large item (e.g., a movie), the item is divided into many packets. If an item is short enough, the entire item can fit into a single packet. Because devices take turns, a single-packet message can be sent without waiting for a long transfer to complete. From a user’s point of view, a short message appears to “sneak in” and use the network while a long transfer continues.

9.6 Each Packet Contains Extra Information

Each packet sent across a network originates at one device and is destined for another. If devices take turns sharing, how can the network hardware tell which device should receive a given packet? The answer is that computer networks operate the same

98 Packet Switching Chap. 9

Because running connections between every pair of communicating devices is prohibitively expensive, networks arrange for multiple com- munications to share a given transmission path.

9.3 Sharing By Taking Turns

How can multiple computers share a transmission path? They take turns. A good analogy comes from early telephone systems. Subscribers who wanted to lower costs could choose a party-line service. Instead of running a separate set of wires to each subscriber’s house, a party-line service meant one set of wires ran down the street and multiple telephones connected to the set of wires. If one of the subscribers on the party line was using the phone, another user on the party line who picked up their phone would hear the conversation. Etiquette dictated that if a subscriber heard someone else talking, they were to hang up immediately and try again later.

Many computer networks use the party-line approach — multiple computers share a transmission system. When a computer has data to send, the computer sends immedi- ately if the network is idle, and must wait for the transfer to complete if the network is in use. The scheme is used for wireless networks, such as Wi-Fi, as well as wired net- works. The point is:

Only one data transfer can occur on a transmission path at a given time. When computers share a network, they take turns sending data.

9.4 Avoiding Long Delays

We have all experienced the downside of taking turns to share a resource: delay. For example, consider an office where employees share a photocopier. Imagine the frustration of needing to copy one page and finding that you must wait for someone who is copying a 900-page document.

If computer networks followed the same approach, long transfers would leave users frustrated. For example, suppose all the subscribers on your street shared a connection that led to your ISP, and suppose all the subscribers took turns using the connection. Imagine how frustrating it would be if you had to wait twenty minutes while a neighbor streamed an HD movie before you could use the connection.

To avoid situations where the data transfer by one device leaves others waiting, networking researchers invented a system that prevents long delays. The idea is straightforward: instead of allowing a given device to use the network for an arbitrarily long time, limit the amount of data that a device can transfer on each turn. The idea, which was invented in the 1960s, is called packet switching, and the unit of data that can be transferred is called a packet. Figure 9.1 illustrates how devices use packet switching.

Sec. 9.4 Avoiding Long Delays 99

The figure shows four devices attached to a network. Device A is sending data to Device C, while Device B sends data to Device D. Each sender divides its outgoing message into packets, and they take turns sending packets. First A sends a packet, then B sends a packet, then A sends a packet, and so on.

A

B

C

D

CDCDC

1st packet from A to C 1st packet from B to D

2nd packet from A to C 2nd packet from B to D 3rd packet from A to C

Shared Network System

Figure 9.1 An illustration of devices taking turns sending packets across a network. Device A communicates with device C, while device B communicates with device D.

9.5 Long Messages And Short Packets

Both LANs and WANs use packet switching, as do both wired and wireless net- works. Furthermore, packet switching is used with all applications. Whether you send a text message, photo, document, video clip, or download a song, the data is always sent in packets. When a user requests a data transfer, software on the sending device divides the item to be sent into packets before sending; on the receiving device, software collects the series of incoming packets, and reconstructs the original item. A user remains unaware that packets are being used.

The maximum size of a packet is set by a network designer. For Ethernet, the maximum size is 1500 bytes of data; for Wi-Fi, 2304 bytes of data can be sent in a sin- gle packet. The details are not important, but remember that when you transfer a large item (e.g., a movie), the item is divided into many packets. If an item is short enough, the entire item can fit into a single packet. Because devices take turns, a single-packet message can be sent without waiting for a long transfer to complete. From a user’s point of view, a short message appears to “sneak in” and use the network while a long transfer continues.

9.6 Each Packet Contains Extra Information

Each packet sent across a network originates at one device and is destined for another. If devices take turns sharing, how can the network hardware tell which device should receive a given packet? The answer is that computer networks operate the same

K392894.indd 127 03/08/18 12:07 pmComer, Douglas E.. The Internet Book : Everything You Need to Know about Computer Networking and How the Internet Works, CRC Press LLC, 2018. ProQuest Ebook Central, http://ebookcentral.proquest.com/lib/apus/detail.action?docID=5502790. Created from apus on 2025-02-06 02:33:15.

C op

yr ig

ht ©

2 01

8. C

R C

P re

ss L

LC . A

ll rig

ht s

re se

rv ed

.

100 Packet Switching Chap. 9

way as a postal system: in addition to the data being sent, each packet contains extra in- formation that specifies the device to which the packet is being sent and the device that sent the packet.

We use the term packet header (or just header) to refer to the part of the packet that holds the extra information. The term was chosen because the header precedes the data. That is, each packet starts with a header and ends with the data being sent. Think of the header as a label that specifies two things: the device that sent the packet and the device that should receive the packet.

9.7 Devices Have Addresses

How is a device identified? Each device on a network is assigned a unique number known as the device’s address. One popular addressing scheme uses the term MAC address. To identify the pair of communicating devices, the header at the begin- ning of a packet contains two important MAC addresses: the MAC address of the de- vice that sent the packet and the MAC address of the device to which the packet is sent. The sender’s address is called the source address, and the receiver’s address is called the destination address. When they travel in a packet, the addresses are actually binary numbers. However, humans who manage networks use more convenient forms (e.g., decimal numbers) to represent addresses. The important idea is:

Each device attached to a network is assigned a unique number called its address. In addition to the data being sent, a packet contains the address of the device that sent the packet and the address of the de- vice to which the packet is sent.

9.8 Packet Size

How large is a packet? The packets used with modern networks usually hold fif- teen hundred characters (bytes) of data. If you download a movie, packet after packet will each contain fifteen hundred bytes of data.

Although packet switching technologies limit the amount of data in a packet, they allow the sender to transmit any size packet up to the maximum. For example, suppose you are engaged in a chat session and send a short reply, ‘OK, it’s a date.” The entire reply takes sixteen characters. So, the chat application will send a single packet that carries sixteen bytes of data. Similarly, when you use a network to transfer a large file or a video, the final packet of the transfer will not be full (unless you happen to transfer a file that is an exact multiple of the packet size).

Sec. 9.9 To Humans, Packet Transmission Seems Instantaneous 101

9.9 To Humans, Packet Transmission Seems Instantaneous

When thinking about packets traveling across computers networks, we must remember that network hardware operates incredibly fast. For example, sending a fifteen-hundred-byte packet across an inexpensive Ethernet LAN takes approximately

0.000012 seconds = 12 millionths of a second (12 microseconds).

Events measured in millionths of a second are hard to imagine. To put it another way, it takes less than two tenths of a second to transfer ten thousand completely full packets across an Ethernet. To summarize:

To avoid long delays, a packet switching system divides each transfer into small packets and arranges for the devices that share a network to take turns sending packets. The time required to send a packet is measured in millionths of a second.

9.10 Sharing Occurs On Demand

Imagine a network with one hundred devices attached. What happens if only one device has packets ready to send? After sending a packet, does the network hardware check each of the ninety-nine computers, find they have nothing to send, and then allow the first computer to send another packet? No. In most networks, only devices that have something ready to send take turns sending. If only one device has packets to send, the device can use the network continuously. If two devices have packets ready to send, they alternate.

The sharing scheme allows a device to participate or stop participating at any time. Once a device has sent its last packet, the device stops using the network. Thus, at any time a device receives an equal share of the network with all other devices that are us- ing the network. If only two devices are actively sharing a network, each will send one-half of the total packets. If three devices are actively sending packets, each device will send one-third of the packets, and so on.

Network sharing is completely automatic because network hardware handles all the details. The hardware does not need to know how many devices are using the network simultaneously, but instead uses a method that allows all active devices to “contend” for access. The key point is:

A packet switching system allows devices to start or stop sending packets at any time. Each device that has packets to send receives a “fair share” of network resources because the hardware is arranged so that the devices with packets to send take turns sending packets.

K392894.indd 128 03/08/18 12:07 pmComer, Douglas E.. The Internet Book : Everything You Need to Know about Computer Networking and How the Internet Works, CRC Press LLC, 2018. ProQuest Ebook Central, http://ebookcentral.proquest.com/lib/apus/detail.action?docID=5502790. Created from apus on 2025-02-06 02:33:15.

C op

yr ig

ht ©

2 01

8. C

R C

P re

ss L

LC . A

ll rig

ht s

re se

rv ed

.

100 Packet Switching Chap. 9

way as a postal system: in addition to the data being sent, each packet contains extra in- formation that specifies the device to which the packet is being sent and the device that sent the packet.

We use the term packet header (or just header) to refer to the part of the packet that holds the extra information. The term was chosen because the header precedes the data. That is, each packet starts with a header and ends with the data being sent. Think of the header as a label that specifies two things: the device that sent the packet and the device that should receive the packet.

9.7 Devices Have Addresses

How is a device identified? Each device on a network is assigned a unique number known as the device’s address. One popular addressing scheme uses the term MAC address. To identify the pair of communicating devices, the header at the begin- ning of a packet contains two important MAC addresses: the MAC address of the de- vice that sent the packet and the MAC address of the device to which the packet is sent. The sender’s address is called the source address, and the receiver’s address is called the destination address. When they travel in a packet, the addresses are actually binary numbers. However, humans who manage networks use more convenient forms (e.g., decimal numbers) to represent addresses. The important idea is:

Each device attached to a network is assigned a unique number called its address. In addition to the data being sent, a packet contains the address of the device that sent the packet and the address of the de- vice to which the packet is sent.

9.8 Packet Size

How large is a packet? The packets used with modern networks usually hold fif- teen hundred characters (bytes) of data. If you download a movie, packet after packet will each contain fifteen hundred bytes of data.

Although packet switching technologies limit the amount of data in a packet, they allow the sender to transmit any size packet up to the maximum. For example, suppose you are engaged in a chat session and send a short reply, ‘OK, it’s a date.” The entire reply takes sixteen characters. So, the chat application will send a single packet that carries sixteen bytes of data. Similarly, when you use a network to transfer a large file or a video, the final packet of the transfer will not be full (unless you happen to transfer a file that is an exact multiple of the packet size).

Sec. 9.9 To Humans, Packet Transmission Seems Instantaneous 101

9.9 To Humans, Packet Transmission Seems Instantaneous

When thinking about packets traveling across computers networks, we must remember that network hardware operates incredibly fast. For example, sending a fifteen-hundred-byte packet across an inexpensive Ethernet LAN takes approximately

0.000012 seconds = 12 millionths of a second (12 microseconds).

Events measured in millionths of a second are hard to imagine. To put it another way, it takes less than two tenths of a second to transfer ten thousand completely full packets across an Ethernet. To summarize:

To avoid long delays, a packet switching system divides each transfer into small packets and arranges for the devices that share a network to take turns sending packets. The time required to send a packet is measured in millionths of a second.

9.10 Sharing Occurs On Demand

Imagine a network with one hundred devices attached. What happens if only one device has packets ready to send? After sending a packet, does the network hardware check each of the ninety-nine computers, find they have nothing to send, and then allow the first computer to send another packet? No. In most networks, only devices that have something ready to send take turns sending. If only one device has packets to send, the device can use the network continuously. If two devices have packets ready to send, they alternate.

The sharing scheme allows a device to participate or stop participating at any time. Once a device has sent its last packet, the device stops using the network. Thus, at any time a device receives an equal share of the network with all other devices that are us- ing the network. If only two devices are actively sharing a network, each will send one-half of the total packets. If three devices are actively sending packets, each device will send one-third of the packets, and so on.

Network sharing is completely automatic because network hardware handles all the details. The hardware does not need to know how many devices are using the network simultaneously, but instead uses a method that allows all active devices to “contend” for access. The key point is:

A packet switching system allows devices to start or stop sending packets at any time. Each device that has packets to send receives a “fair share” of network resources because the hardware is arranged so that the devices with packets to send take turns sending packets.

100 Packet Switching Chap. 9

way as a postal system: in addition to the data being sent, each packet contains extra in- formation that specifies the device to which the packet is being sent and the device that sent the packet.

We use the term packet header (or just header) to refer to the part of the packet that holds the extra information. The term was chosen because the header precedes the data. That is, each packet starts with a header and ends with the data being sent. Think of the header as a label that specifies two things: the device that sent the packet and the device that should receive the packet.

9.7 Devices Have Addresses

How is a device identified? Each device on a network is assigned a unique number known as the device’s address. One popular addressing scheme uses the term MAC address. To identify the pair of communicating devices, the header at the begin- ning of a packet contains two important MAC addresses: the MAC address of the de- vice that sent the packet and the MAC address of the device to which the packet is sent. The sender’s address is called the source address, and the receiver’s address is called the destination address. When they travel in a packet, the addresses are actually binary numbers. However, humans who manage networks use more convenient forms (e.g., decimal numbers) to represent addresses. The important idea is:

Each device attached to a network is assigned a unique number called its address. In addition to the data being sent, a packet contains the address of the device that sent the packet and the address of the de- vice to which the packet is sent.

9.8 Packet Size

How large is a packet? The packets used with modern networks usually hold fif- teen hundred characters (bytes) of data. If you download a movie, packet after packet will each contain fifteen hundred bytes of data.

Although packet switching technologies limit the amount of data in a packet, they allow the sender to transmit any size packet up to the maximum. For example, suppose you are engaged in a chat session and send a short reply, ‘OK, it’s a date.” The entire reply takes sixteen characters. So, the chat application will send a single packet that carries sixteen bytes of data. Similarly, when you use a network to transfer a large file or a video, the final packet of the transfer will not be full (unless you happen to transfer a file that is an exact multiple of the packet size).

Sec. 9.9 To Humans, Packet Transmission Seems Instantaneous 101

9.9 To Humans, Packet Transmission Seems Instantaneous

When thinking about packets traveling across computers networks, we must remember that network hardware operates incredibly fast. For example, sending a fifteen-hundred-byte packet across an inexpensive Ethernet LAN takes approximately

0.000012 seconds = 12 millionths of a second (12 microseconds).

Events measured in millionths of a second are hard to imagine. To put it another way, it takes less than two tenths of a second to transfer ten thousand completely full packets across an Ethernet. To summarize:

To avoid long delays, a packet switching system divides each transfer into small packets and arranges for the devices that share a network to take turns sending packets. The time required to send a packet is measured in millionths of a second.

9.10 Sharing Occurs On Demand

Imagine a network with one hundred devices attached. What happens if only one device has packets ready to send? After sending a packet, does the network hardware check each of the ninety-nine computers, find they have nothing to send, and then allow the first computer to send another packet? No. In most networks, only devices that have something ready to send take turns sending. If only one device has packets to send, the device can use the network continuously. If two devices have packets ready to send, they alternate.

The sharing scheme allows a device to participate or stop participating at any time. Once a device has sent its last packet, the device stops using the network. Thus, at any time a device receives an equal share of the network with all other devices that are us- ing the network. If only two devices are actively sharing a network, each will send one-half of the total packets. If three devices are actively sending packets, each device will send one-third of the packets, and so on.

Network sharing is completely automatic because network hardware handles all the details. The hardware does not need to know how many devices are using the network simultaneously, but instead uses a method that allows all active devices to “contend” for access. The key point is:

A packet switching system allows devices to start or stop sending packets at any time. Each device that has packets to send receives a “fair share” of network resources because the hardware is arranged so that the devices with packets to send take turns sending packets.

K392894.indd 129 03/08/18 12:07 pmComer, Douglas E.. The Internet Book : Everything You Need to Know about Computer Networking and How the Internet Works, CRC Press LLC, 2018. ProQuest Ebook Central, http://ebookcentral.proquest.com/lib/apus/detail.action?docID=5502790. Created from apus on 2025-02-06 02:33:15.

C op

yr ig

ht ©

2 01

8. C

R C

P re

ss L

LC . A

ll rig

ht s

re se

rv ed

.

102 Packet Switching Chap. 9

9.11 Relevance To The Internet

Like most computer networks, the Internet is a packet switching system. Packet switching allows many communications to proceed across the Internet simultaneously without requiring one user to wait for another user to finish their communication. As a consequence of the technology, whenever a user transfers a data item across the Inter- net, network software on the sending machine must divide the data into packets, and network software on the receiving machine must reconstruct packets to produce the ori- ginal item. For example, a photo must be divided into packets for transfer across the Internet, and then reassembled into a copy of the photo at the receiving side. To sum- marize:

All data is transferred across the Internet in packets. A sender divides a message or document into packets and transfers the packets across the Internet. A receiver reassembles the original message from the packets that arrive. Packets from many machines traverse the In- ternet at the same time.

9.12 Summary

The fundamental technique that computer networks use to ensure fair access to shared network resources is known as packet switching. Before data can be transferred across a packet switching network, the data must be divided into individual packets. A typical packet can hold up to fifteen hundred characters (bytes) of data.

Every device is assigned a unique number known as the device’s address. Each packet contains a header that specifies the address of the device that sent the packet and the address of the device to which the packet is sent. Devices that share access to a net- work take turns sending packets. On each turn, a given device can send one packet.

EXERCISES

9.1 If you have DSL or cable Internet service, look at the label on the bottom of your DSL or cable modem and find its MAC address. Hint: a MAC address consists of twelve charac- ters, including digits 0 through 9 and letters A through F.

9.2 Sue and Paul are sitting in a coffee shop using the Internet when suddenly everything stops working. After ten seconds, Paul says that he heard that devices on the Internet use pack- ets, and they are probably blocked waiting because someone else is sending a packet. Can Paul be correct? Explain.

9.3 Think about a coffee shop. Name five types of devices (either that customers have or the coffee shop has) that use the Internet.

Exercises 103

9.4 Suppose that every individual on earth has both a laptop and a smart phone, and suppose each device needs a unique address. How many total addresses would be needed? (Hint: search online to find the world’s population.)

9.5 On a police drama when a criminal is accessing a victim’s computer, the script has the po- lice capture data and use the data to identify the criminal’s computer. What information in a packet can be used to identify the sending computer?

K392894.indd 130 03/08/18 12:07 pmComer, Douglas E.. The Internet Book : Everything You Need to Know about Computer Networking and How the Internet Works, CRC Press LLC, 2018. ProQuest Ebook Central, http://ebookcentral.proquest.com/lib/apus/detail.action?docID=5502790. Created from apus on 2025-02-06 02:33:15.

C op

yr ig

ht ©

2 01

8. C

R C

P re

ss L

LC . A

ll rig

ht s

re se

rv ed

.

102 Packet Switching Chap. 9

9.11 Relevance To The Internet

Like most computer networks, the Internet is a packet switching system. Packet switching allows many communications to proceed across the Internet simultaneously without requiring one user to wait for another user to finish their communication. As a consequence of the technology, whenever a user transfers a data item across the Inter- net, network software on the sending machine must divide the data into packets, and network software on the receiving machine must reconstruct packets to produce the ori- ginal item. For example, a photo must be divided into packets for transfer across the Internet, and then reassembled into a copy of the photo at the receiving side. To sum- marize:

All data is transferred across the Internet in packets. A sender divides a message or document into packets and transfers the packets across the Internet. A receiver reassembles the original message from the packets that arrive. Packets from many machines traverse the In- ternet at the same time.

9.12 Summary

The fundamental technique that computer networks use to ensure fair access to shared network resources is known as packet switching. Before data can be transferred across a packet switching network, the data must be divided into individual packets. A typical packet can hold up to fifteen hundred characters (bytes) of data.

Every device is assigned a unique number known as the device’s address. Each packet contains a header that specifies the address of the device that sent the packet and the address of the device to which the packet is sent. Devices that share access to a net- work take turns sending packets. On each turn, a given device can send one packet.

EXERCISES

9.1 If you have DSL or cable Internet service, look at the label on the bottom of your DSL or cable modem and find its MAC address. Hint: a MAC address consists of twelve charac- ters, including digits 0 through 9 and letters A through F.

9.2 Sue and Paul are sitting in a coffee shop using the Internet when suddenly everything stops working. After ten seconds, Paul says that he heard that devices on the Internet use pack- ets, and they are probably blocked waiting because someone else is sending a packet. Can Paul be correct? Explain.

9.3 Think about a coffee shop. Name five types of devices (either that customers have or the coffee shop has) that use the Internet.

Exercises 103

9.4 Suppose that every individual on earth has both a laptop and a smart phone, and suppose each device needs a unique address. How many total addresses would be needed? (Hint: search online to find the world’s population.)

9.5 On a police drama when a criminal is accessing a victim’s computer, the script has the po- lice capture data and use the data to identify the criminal’s computer. What information in a packet can be used to identify the sending computer?

102 Packet Switching Chap. 9

9.11 Relevance To The Internet

Like most computer networks, the Internet is a packet switching system. Packet switching allows many communications to proceed across the Internet simultaneously without requiring one user to wait for another user to finish their communication. As a consequence of the technology, whenever a user transfers a data item across the Inter- net, network software on the sending machine must divide the data into packets, and network software on the receiving machine must reconstruct packets to produce the ori- ginal item. For example, a photo must be divided into packets for transfer across the Internet, and then reassembled into a copy of the photo at the receiving side. To sum- marize:

All data is transferred across the Internet in packets. A sender divides a message or document into packets and transfers the packets across the Internet. A receiver reassembles the original message from the packets that arrive. Packets from many machines traverse the In- ternet at the same time.

9.12 Summary

The fundamental technique that computer networks use to ensure fair access to shared network resources is known as packet switching. Before data can be transferred across a packet switching network, the data must be divided into individual packets. A typical packet can hold up to fifteen hundred characters (bytes) of data.

Every device is assigned a unique number known as the device’s address. Each packet contains a header that specifies the address of the device that sent the packet and the address of the device to which the packet is sent. Devices that share access to a net- work take turns sending packets. On each turn, a given device can send one packet.

EXERCISES

9.1 If you have DSL or cable Internet service, look at the label on the bottom of your DSL or cable modem and find its MAC address. Hint: a MAC address consists of twelve charac- ters, including digits 0 through 9 and letters A through F.

9.2 Sue and Paul are sitting in a coffee shop using the Internet when suddenly everything stops working. After ten seconds, Paul says that he heard that devices on the Internet use pack- ets, and they are probably blocked waiting because someone else is sending a packet. Can Paul be correct? Explain.

9.3 Think about a coffee shop. Name five types of devices (either that customers have or the coffee shop has) that use the Internet.

Exercises 103

9.4 Suppose that every individual on earth has both a laptop and a smart phone, and suppose each device needs a unique address. How many total addresses would be needed? (Hint: search online to find the world’s population.)

9.5 On a police drama when a criminal is accessing a victim’s computer, the script has the po- lice capture data and use the data to identify the criminal’s computer. What information in a packet can be used to identify the sending computer?

K392894.indd 131 03/08/18 12:07 pmComer, Douglas E.. The Internet Book : Everything You Need to Know about Computer Networking and How the Internet Works, CRC Press LLC, 2018. ProQuest Ebook Central, http://ebookcentral.proquest.com/lib/apus/detail.action?docID=5502790. Created from apus on 2025-02-06 02:33:15.

C op

yr ig

ht ©

2 01

8. C

R C

P re

ss L

LC . A

ll rig

ht s

re se

rv ed

.

Chapter Contents 10 Internet: A Network Of Networks 10.1 Introduction 107 10.2 Building A Global Network 107 10.3 Two Fundamental Concepts 108 10.4 Using A Specialized Computer To Interconnect Networks 109 10.5 Internet Terminology: Routers And Hosts 110 10.6 Building A Large Virtual Network 111 10.7 The Internet Includes Multiple Types Of Networks 113 10.8 Ownership, ISPs, And Transit Traffic 113 10.9 A Hierarchy Of ISPs 114 10.10 Peering Arrangements At The Center Of The Internet 115 10.11 An Example Trip Through The Internet 116 10.12 The Internet Approach Revolutionized Networking 116 10.13 Summary 117

K392894.indd 132 03/08/18 12:07 pmComer, Douglas E.. The Internet Book : Everything You Need to Know about Computer Networking and How the Internet Works, CRC Press LLC, 2018. ProQuest Ebook Central, http://ebookcentral.proquest.com/lib/apus/detail.action?docID=5502790. Created from apus on 2025-02-06 02:33:15.

C op

yr ig

ht ©

2 01

8. C

R C

P re

ss L

LC . A

ll rig

ht s

re se

rv ed

.

Chapter Contents 10 Internet: A Network Of Networks 10.1 Introduction 107 10.2 Building A Global Network 107 10.3 Two Fundamental Concepts 108 10.4 Using A Specialized Computer To Interconnect Networks 109 10.5 Internet Terminology: Routers And Hosts 110 10.6 Building A Large Virtual Network 111 10.7 The Internet Includes Multiple Types Of Networks 113 10.8 Ownership, ISPs, And Transit Traffic 113 10.9 A Hierarchy Of ISPs 114 10.10 Peering Arrangements At The Center Of The Internet 115 10.11 An Example Trip Through The Internet 116 10.12 The Internet Approach Revolutionized Networking 116 10.13 Summary 117

Chapter Contents 10 Internet: A Network Of Networks 10.1 Introduction 107 10.2 Building A Global Network 107 10.3 Two Fundamental Concepts 108 10.4 Using A Specialized Computer To Interconnect Networks 109 10.5 Internet Terminology: Routers And Hosts 110 10.6 Building A Large Virtual Network 111 10.7 The Internet Includes Multiple Types Of Networks 113 10.8 Ownership, ISPs, And Transit Traffic 113 10.9 A Hierarchy Of ISPs 114 10.10 Peering Arrangements At The Center Of The Internet 115 10.11 An Example Trip Through The Internet 116 10.12 The Internet Approach Revolutionized Networking 116 10.13 Summary 117

K392894.indd 133 03/08/18 12:07 pmComer, Douglas E.. The Internet Book : Everything You Need to Know about Computer Networking and How the Internet Works, CRC Press LLC, 2018. ProQuest Ebook Central, http://ebookcentral.proquest.com/lib/apus/detail.action?docID=5502790. Created from apus on 2025-02-06 02:33:15.

C op

yr ig

ht ©

2 01

8. C

R C

P re

ss L

LC . A

ll rig

ht s

re se

rv ed

.

10

Internet: A Network Of Networks

10.1 Introduction

The previous chapter describes packet switching and shows why dividing long messages into packets allows computers to share a transmission path without introduc- ing arbitrarily long delays. This chapter describes how multiple packet switching net- works can be interconnected to form an Internet that functions like a single, large net- work.

10.2 Building A Global Network

How can we build a global communication system? As Chapter 7 points out, no single technology suffices for all purposes. Many packet switching technologies exist because each has been designed to meet constraints of speed, distance, and cost. So, the question becomes: can we allow groups to choose a network technology that meets the group’s needs, but still be able to connect multiple networks together? Unfortunately the simplest approach — connect the wires from one network to the wires of another — doesn’t work. To understand why, one must know that designers do not make all tech- nologies compatible. Details, such as the electrical voltages used and the maximum size of a packet often differ. In fact, directly connecting wires can permanently damage the hardware. Consequently, when they designed the Internet, researchers did not envi- sion merely connecting wires of incompatible networks.

107

K392894.indd 134 03/08/18 12:07 pmComer, Douglas E.. The Internet Book : Everything You Need to Know about Computer Networking and How the Internet Works, CRC Press LLC, 2018. ProQuest Ebook Central, http://ebookcentral.proquest.com/lib/apus/detail.action?docID=5502790. Created from apus on 2025-02-06 02:33:15.

C op

yr ig

ht ©

2 01

8. C

R C

P re

ss L

LC . A

ll rig

ht s

re se

rv ed

.