Instant Download with all chapters and Answers
Sample Chapters
*you will get solution manuals in PDF in best viewable format after buy*
Chapter 2
Network Standards
Standards Govern Message Exchanges
Network Standards (Protocols)
Message Order, Semantics, and Syntax
Test Your Understanding
1. a) Give the definition of network standards that this chapter introduced.
Network standards govern the exchange of messages between hardware or
software processes on different host computers, including message order,
semantics, syntax, reliability, and connection orientation.
b) What three things about message exchanges do network standards govern?
Message order, semantics, and syntax
c) Give an example not involving networking in which the order in which you do things
can make a big difference.
Answers will vary
Example: Installing a printer on a computer (when to power it on, etc.)
d) Distinguish between syntax and semantics.
Syntax governs the organization of messages.
Semantics defines the meaning of messages.
Syntax: General Message Organization
Test Your Understanding
2. a) What are the three general parts of messages?
The three general parts of messages are the header, the data field, and the trailer.
b) What does the data field contain?
The data field contains the content delivered by the message.
c) What is the definition of a header?
The header is everything that comes before the data field.
d) Is there always a data field in a message?
No, there is not always a data field in a message.
e) What is the definition of a trailer?
The trailer is everything that comes after the data field.
f) Are trailers common?
No, trailers are not common. <When a trailer exists, it is usually at the data link
layer, and not even all data link standards use them.>
g) Distinguish between headers and header fields.
The header is everything that comes before the data field. A header field is a
subdivision of the header.
Reliability and Connections
Reliability
Test Your Understanding
3. a) What is reliability?
Reliability is the condition wherein errors are corrected by resending lost or
damaged messages.
b) How does TCP implement reliability?
TCP implements reliability with its use of acknowledgements (ACKs). The
receiver acknowledges every correctly received TCP segment. The original sender
retransmits any segments that are not acknowledged.
c) In TCP, what is the receiver‘s role in reliability?
In TCP, the receiver‘s role in reliability is to send back an ACK segment for every
correct TCP segment it receives.
d) In TCP, what is the sender‘s role in reliability?
In TCP, the sender‘s role in reliability is to resend a segment for which it does not
receive an acknowledgement.
e) What is the disadvantage of reliability?
It places a heavy load on computers.
NOT FOR DISTRIBUTION TO STUDENTS
Connection-Oriented and Connectionless Protocols
Connectionless and Unreliable Protocols Dominate
Test Your Understanding
4. a) Distinguish between connectionless and connection-oriented protocols.
Connection-oriented services have explicit openings and closings. In contrast,
connectionless services do not establish connections before transmitting or close
the connection after they have finished transmitting.
b) Which can have sequence numbers?
Connection-oriented service can have sequence numbers.
c) What are the advantages that sequence numbers bring to connection-oriented
protocols?
Thanks to sequence numbers, the parties can tell when a message is lost. (There
will be a gap in the sequence numbers.)
Acknowledgements can refer to specific messages according to the sequence
numbers of these messages.
Long messages can be fragmented into many smaller messages that can fit inside
of packets. The fragments will be given sequence numbers so that they can be
assembled at the other end. Fragmentation followed by reassembly is an
important concept in networking.
Messages can refer to an earlier message by sequence number. This is important
in database-based transaction processes where several messages must be
exchanged to make a purchase, record a transaction, or do some other common
business task.
d) Explain fragmentation and reassembly.
The sender divides a message into a number of fragments small enough to fit into
individual packets.
Each fragment is given a sequence number.
The receiver reassembles the fragments by sequence number into the original
message.
e) What is the disadvantage of connection-oriented protocols?
They are expensive, placing a heavy load on networks and computers.
f) Are most protocols connectionless or connection-oriented?
Most are connectionless.
g) Are most protocols reliable or unreliable?
Most are unreliable
Layered Standards Architectures
Architectures
Test Your Understanding
5. a) What is a network architecture?
A network architecture is a broad plan that specifies everything necessary for two
application programs on different networks on an internet to be able to work
together effectively.
b) What is the most popular network architecture today?
The most popular standards architecture for networking today is the Hybrid
TCP/IP–OSI Architecture.
c) In layered standards architectures, to what layer or layers does a layer provide service?
In layered standards architecture, each layer provides services to the layer above
it.
6. Why do standards architectures break down the standards development process into
layers?
First, it is a good strategy to break up a major task into individual, more
manageable, pieces.
Second, breaking down the standards development process into layers allows
team members to be assigned individual tasks that suit their skills—for example
having specialists in a particular application create application layer standards
without having to worry about physical layer standards.
Third, layering means that development at one layer is freed of concerns at other
layers.
Fourth, layering allows standards to be updated or changed at various layers
independently.
Layer 1 and Layer 2 Standards for Switched Networks (Switched
LANs and WANs)
Physical Links
Test Your Understanding
7. a) What devices does a physical link connect?
A physical link connects adjacent devices connected by a transmission medium—
a computer to a switch, a switch to a switch, or a switch to a router.
b) What is a data link?
A data link is the path that a frame takes across a single switched network.
c) Five switches separate two computers on a switched network. How many physical
links are there between the two computers?
There are six physical links between the two computers.
d) How many data links are there between them?
There is one data link between them.
e) What do data link layer standards govern?
Data link layer standards govern the transmission of frames across a single
switched network—typically by sending them through several switches along the
data link. Data link standards also govern frame organization, reliability, and
other matters.
f) Which layers govern switched LAN transmission?
The physical and data link layers govern switched LAN transmission. (A switched
LAN is a single switched network.) Students need to do some thinking here. They
know from Chapter 1 that a switched LAN is a single switched network. They
know from this chapter that single switched networks involve the physical and
data link layers and only these layers.
g) Which layers govern switched WAN transmission?
The physical and data link layers govern switched WAN transmission.
Standards for Routed Network Transmission (Layers 3 and 4)
Test Your Understanding
8. a) What do the internet and transport layers do collectively?
Collectively, the internet and transport layers govern the communication between
two end hosts across an internet.
b) Distinguish between what the internet and transport layer standards govern.
Internet layer standards govern the transmission of packets across an internet—
typically by sending them through several routers along the route. Internet layer
standards also govern packet organization and other matters.
Transport layer standards govern the aspects of end-to-end communication
between two end hosts that are not handled by the internet layer.
c) What is the main internet layer standard?
The Internet Protocol (IP).
d) What errors does the transport layer usually fix?
The transport layer usually fixes all errors created at the transport layer or lower
layers.
e) What does it mean in this book if internet is spelled with a lower-case i?
It refers to a single routed network or to the internet layer.
Test Your Understanding
9. a) What do application layer standards govern?
Application layer standards govern how two applications work with each other,
even if they are from different vendors.
b) Which layer has the most standards? Why is this the case?
The application layer has the most standards.
This is the case because there are many different applications, each with its own
standard.
Layers 1 (Physical) and 2 (Data Link) in Ethernet
Ethernet Physical Layer Standards
Test Your Understanding
10. What does the sending physical layer process do with the bits of the frame?
It translates them into signals.
Ethernet Frames
Test Your Understanding
11. What is an octet?
An octet is a collection of eight bits. <This is the same as a byte.>
12. a) How many bits long are Ethernet addresses?
Ethernet addresses are 48 bits long.
b) When are Ethernet addresses set on NICs?
Ethernet addresses are set on NICs at the factory before the card ships.
c) In what notation are Ethernet addresses typically written for human reading?
Ethernet addresses are typically written in hexadecimal notation for human
reading.
d) What device in an Ethernet network besides the destination host reads the Ethernet
address?
Switches between the source and destination hosts read the Ethernet address.
e) What is its purpose in reading the Ethernet address?
The switch‘s purpose in doing so is to look up the Ethernet address in the
switching table so that the switch can decide which port number to select to send
the frame back out.
13. Where is the IP packet carried in an Ethernet frame?
The Ethernet data field usually contains an IP packet.
14. a) How many bits long is the Ethernet frame check sequence field?
The Ethernet frame check sequence field is 32 bits long.
b) What is the purpose of the Ethernet frame check sequence field?
Its purpose is to allow the receiver to detect whether errors have occurred during
transmission.
c) How does the receiving NIC use the value in the frame check sequence field?
The receiving NIC recomputes the frame check sequence number and compares
its result with the value contained in the frame check sequence field. If the two do
not match, the receiving NIC knows that an error must have occurred during
transmission.
d) What happens if a receiving NIC detects an error?
If a receiving NIC detects an error, the NIC‘s data link layer process simply
discards the frame. There is no request for retransmission.
e) Does this error detection and discarding process make Ethernet a reliable standard?
Explain.
This error detection and discarding process does not make Ethernet a reliable
standard, because there is no error correction.
Ethernet Is Unreliable and Connectionless
Test Your Understanding
15. a) Is Ethernet connectionless or connection-oriented? Explain.
Ethernet is connectionless. NICs send Ethernet frames without opens, closes,
acknowledgements, or sequence numbers.
b) Is Ethernet reliable or unreliable? Explain.
Ethernet is unreliable. Errors are detected, but they are not corrected.
Layer 3: The Internet Protocol (IP)
Layer 2 versus Layer 3
Test Your Understanding
16. a) Four switched networks are involved in transmissions from the source to the
destination host. How many packets will there be along the way when the source host
transmits a packet?
There will be one packet.
b) How many frames will there be along the way?
There will be four frames.
c) How many routes will there be along the way?
There will be one route.
d) How many data links will there be along the way?
There will be four data links.
e) How many destination IP addresses will there be?
There will be one destination IP address (that of the destination host).
f) How many data link layer destination addresses will there be?
There will be four data link layer destination addresses.
g) In the data link layer destination address field of the frame in the first switched
network, the destination address is the destination address of which device?
The data link layer address of the router that connects the first and second
switched networks will be the Layer 2 data link layer destination address of the
frame. <Actually, the data link layer address of the port into which the frame
arrives. Each port on the router has a different data link layer address.>
h) What will be the destination IP address of the packet contained in that frame?
The IP address of the destination host.
The IP Packet
Test Your Understanding
17. a) How many octets long is an IP header if there are no options? (Look at Figure 2-12.)
If there are no options, the IP header will be 20 octets.
b) What is the bit number of the first bit in the destination address field? (Remember that
the first bit in binary counting is Bit 0.)
128. <The first bit on each line is 0, 32, 64, 96, and 128.>
c) How long are IP addresses?
IP addresses are 32 bits long.
d) You have two addresses: B7-23-DD-6F-C8-AB and 217.42.18.248. Specify what kind
of address each address is.
B7-23-DD-6F-C8-AB is an Ethernet address.
217.42.18.248 is an IP address.
e) What device in an internet besides the destination host reads the destination IP
address?
Each router along the way reads the destination IP address.
f) What is this device‘s purpose in doing so?
The router reads the IP address in order to learn how to forward the IP packet to
the next router or to the destination host itself.
IP Characteristics
Test Your Understanding
18. a) Is IP connectionless or connection-oriented?
IP is connectionless.
b) Is IP reliable or unreliable?
IP is unreliable.
Layer 4: The Transport Layer
Layers 3 and 4
Test Your Understanding
19. Ten routers separate two hosts. a) How many internet layer processes will be active on
the two hosts and the routers between them?
There will be 12, one on each router and host.
b) How many transport layer processes will be active? (The answer is not directly in the
book. You will have to think about this one a little.)
Two will be active: one each on the source and the destination hosts.
c) Which layers are hop-by-hop layers? (The physical layer is not considered to be either
a hop-by-hop or an end-to-end layer.)
The internet and data link layers are hop-by-hop layers.
d) Which layers are end-to-end layers? (The physical layer is not considered to be either a
hop-by-hop or an end-to-end layer.)
The application and transport layers are end-to-end layers.
TCP: A Reliable Protocol
Test Your Understanding
20. a) Why are most standards unreliable?
Most standards are unreliable because reliability is expensive. It takes a great deal
of processing power to test for errors and to implement retransmission. It also
places a heavy burden on internet traffic because of acknowledgements and the
other supervisory traffic needed to implement reliability.
b) For what two reasons is making TCP reliable a good choice?
Making TCP reliable is a good choice because TCP operates at the highest level
below the application layer, so TCP can provide error-free data to the application
program without each lower layer having to do error correction.
In addition, TCP is an end-to-end layer, so error correction only has to be done on
the two hosts.
The User Datagram Protocol (UDP)
Test Your Understanding
21. Compare TCP and UDP in terms of layer of operation, connection-orientation, reliability,
and burden (traffic and processing on devices).
Both protocols work at the transport layer. While TCP is connection-oriented and
reliable, so it is burdensome on the two hosts and on the network, UDP is
connectionless and unreliable, so it places a lighter load on the two hosts and the
network.
The Only Protocols at the Transport Layer
Test Your Understanding
22. At the transport layer, what are the only TCP/IP protocols?
TCP and UDP.
Layer 5: HTTP and Other Application Standards
Test Your Understanding
23. a) Is the application layer standard always HTTP?
No, the application layer standard is not always HTTP.
b) Which layer has the most standards?
The standard at the application layer standard is always HTTP.
c) At which layer would you find standards for instant messaging? (The answer is not
explicitly in this section.)
Instant messaging is an application, so one would expect to find standards for
instant messaging at the application layer.
Vertical Communication on Hosts, Switches, and
Routers
Layered Communication on the Source Host
Test Your Understanding
24. a) When a layer creates a message, what does it usually do immediately afterward?
When a layer creates a message, it usually passes the message down to the next-
lower layer immediately.
b) What does the layer below it usually do after receiving the next-higher-layer message?
After receiving the next-higher-layer message, a layer usually encapsulates the
received message in the data field of its own message. It then adds a header and,
sometimes, a trailer.
<Why ―usually?‖ First, TCP may first have to fragment the application message
into several TCP messages (segments).>
<Second, as we will see in Chapter 8, when TCP receives the first application
message to a destination host, what it does depends on whether there is a
connection open to the TCP process on the destination host. If there is a
connection open, the TCP process will encapsulate the application message and
pass it on immediately. However, if no connection is open, the TCP process will
open a connection. After it opens a connection to that host and application, it will
deliver the message.>
c) What is encapsulation?
Encapsulation is placing a message in the data field of another message.
d) With ̳Web communication using HTTP, what message does IP encapsulate in packet
data fields?
IP encapsulates the TCP segment (alternatively, a TCP header and HTTP
message).
25. a) What are the two steps after a layer process creates its layer message?
When a layer N creates its message, the layer process passes the message down to
the next-lower (N – 1) layer.
The N – 1 layer encapsulates the message in the data field of the Layer N – 1
message and adds a Layer N – 1 header and (at the data link layer only) perhaps a
layer N – 1 trailer. The Layer N – 1 process repeats the cycle.
b) What is the final frame if SMTP (an e-mail protocol that requires TCP) is used at the
application layer and if Frame Relay (which has a header and a trailer) is used instead of
Ethernet at the data link layer?
The final frame will be a Frame Relay header, an IP header, a TCP header, an
SMTP message, and a Frame Relay trailer.
c) What is the final frame if SNMP (which requires UDP) is used at the application layer
and if the ATM data link layer protocol (which has a header but no trailer) is used instead
of Ethernet at the data link layer?
The final frame will be an ATM header, an IP header, a UDP header, and an
SNMP message.
On the Destination Host
Test Your Understanding
26. a) Which host decapsulates—the sending host or the receiving host?
The receiving host decapsulates.
b) Describe what each layer‘s process does on the receiving host when the host receives
an Ethernet frame containing an HTTP message.
The physical layer process turns the signals into the bits of the frame and passes
the frame to the data link layer.
The data link layer process checks the Ethernet frame for errors. If there are no
errors, it decapsulates the IP packet and passes the packet up to the internet layer
process.
The internet layer process checks the IP packet header for errors. If there are no
errors, it decapsulates the TCP segment and passes the segment up to the transport
layer process.
The transport layer process checks the TCP segment for errors. If there are no
errors, it decapsulates the HTTP message and passes this message up to the HTTP
application layer process.
On Switches and Routers along the Way
Test Your Understanding
27. a) Why are switches called Layer 2 devices?
Switches are called Layer 2 devices because the highest layer in switching is the
data link layer (Layer 2).
b) Why are routers called Layer 3 devices?
Routers are called Layer 3 devices because the highest layer in routing is the
internet layer (Layer 3).
c) Do routers first encapsulate or decapsulate? (The answer is not explicitly in the text.
Look at Figure 2-17.)
Routers first decapsulate packets from incoming frames, then encapsulate packets
in new frames for sending out another port.
Layering on a Source Router
Test Your Understanding
28. a) Why is there no transport or application content to the final frame in Figure 2-18?
The internet layer initiates the ICMP message transmission. Higher layers are not
involved at all, so there are no higher-layer messages or headers.
b) To create the frame in the figure, at what layer do you begin?
Layer 3
c) At what layer is an ICMP message?
Layer 3
Combining Vertical and Horizontal Communication
Test Your Understanding
29. To what software process is the transport layer message addressed?
To the transport layer process on the destination host.
Major Standards Architectures
TCP/IP and OSI Architectures
Test Your Understanding
30. a) What is a standards architecture?
A standards architecture is a family of related standards that collectively allow an
application program on one machine on an internet to communicate with another
application program on another machine on that same internet.
b) What are the two dominant network standards architectures?
The two dominant network standards architectures are OSI and TCP/IP.
c) Are they competitors?
Although OSI and TCP/IP sometimes are viewed as competitors, they actually
work together in most corporate networks.
Test Your Understanding
31. a) What standards agencies are responsible for the OSI standards architecture? Just give
the acronyms.
The standards agencies responsible for the OSI standards architecture are the
International Organization for Standardization (ISO) and the International
Telecommunications Union–Telecommunications Standards Sector (ITU-T).
b) At which layers do OSI standards dominate usage?
OSI is dominant at the physical and data link layers.
c) Name and describe the functions of OSI Layer 5.
OSI Layer 5 is the OSI session layer. It initiates and maintains a connection
between application programs on different computers. It is especially good for
database applications. If communication fails during a transaction, the entire
transaction does not have to be done over—only the work since the last rollback
point.
d) Name and describe the intended use of OSI Layer 6.
OSI Layer 6 is the OSI presentation layer. It is designed to handle data formatting
differences between two computers, as well as compression and encryption.
e) How is the OSI presentation layer actually used?
The OSI presentation layer is actually used as a category for data file formats.
f) Beginning with the physical layer (Layer 1), give the name and number of the OSI
layers.
1. Physical
2. Data link
3. Network
4. Transport
5. Session
6. Presentation
7. Application
TCP/IP
Test Your Understanding
32. a) Which of the following is an architecture: TCP/IP, TCP, or IP?
TCP/IP is an architecture.
b) Which of the following are standards: TCP/IP, TCP, or IP?
TCP and IP are standards.
c) What is the standards agency for TCP/IP?
The standards agency for TCP/IP is the Internet Engineering Task Force (IETF).
d) Why have this agency‘s standards been so successful?
IETF TCP/IP standards have been successful because they tend to be simple
standards that can be implemented quickly and inexpensively. <Not primarily
because of the use of these standards on the Internet.>
e) What are most of this agency‘s documents called?
Most of this agency‘s documents are called requests for comment (RFCs).
f) At which layers is TCP/IP dominant?
TCP/IP is dominant at the internet and transport layers.
g) How dominant is TCP/IP today at these layers compared with OSI‘s dominance at the
physical and data link layers?
TCP‘s dominance at these layers is not nearly universal, as OSI‘s dominance at
the physical and data link layers is.
The Application Layer
Test Your Understanding
33. a) Is any standards architecture dominant at the application layer?
No standards architecture is dominant at the application layer, although IETF
protocols are widely used.
b) Do almost all applications run over TCP/IP standards at the internet and transport
layers?
Yes.
TCP/IP and OSI: The Hybrid TCP/IP–OSI Standards Architecture
Test Your Understanding
34. a) What layers of the hybrid TCP/IP–OSI standards architecture use OSI standards?
The physical and data link layers of the hybrid TCP/IP–OSI standards architecture
use OSI standards.
b) What layers use TCP/IP standards?
The internet and transport layers primarily use TCP/IP standards.
c) Do switched LAN standards come from OSI or TCP/IP? Explain. (The answer is not
explicitly in this section.)
Switched LAN standards come from OSI because switched LANs are single
switched networks.
d) Do switched WAN standards come from OSI or TCP/IP? Explain. (Again, the answer
is not explicitly in this section.)
Switched WAN standards come from OSI because WANs are single switched
networks even if they are very large.
A Multiprotocol World at Higher Layers
Test Your Understanding
35. a) Under what circumstances might you encounter IPX/SPX standards?
You might encounter IPX/SPX standards if Novell NetWare servers were being
used.
b) SNA standards?
You might encounter SNA standards if IBM mainframe computers were
communicating over the network.
c) AppleTalk standards?
You might encounter AppleTalk standards if Apple Macintoshes were
communicating over the network.
Conclusion
Synopsis
End-of-Chapter Questions
Basic Thought Questions
1. To open a TCP connection to the transport layer process on Host B, the TCP process on
host A sends a TCP SYN segment. What will the final frame be that Host A sends if Host
A is on a Frame Relay (FR) switched network? FR frames have both headers and trailers.
The application layer process is not at all involved in the connection-opening attempt.
FR-H IP-H TCP-H FR-T
Note that there is no TCP message. Start with the TCP header at L4.
Then add the IP header (there is no IP trailer) at L3
Then add the FR header and trailer at L2
2. Figure 2-11 shows the fields in an Ethernet frame. Ethernet is the dominant standard for
switched LANs. However, there are many other data link standards. One example is the
Point-to-Point Protocol (PPP). This protocol is used at the data link layer to connect two
routers with a point-to-point leased line from the telephone company. PPP frames begin
and end with a one-octet flag field containing the content 01111110. These
unambiguously signal the start of a new frame and the end of that frame, respectively.
The second two octets always have the values 11111111 and 00000011. These are the
address and control fields, respectively. They exist for historical reasons that are no
longer important. Obviously, there is no need for an address field in a point-to-point
connection, and the function of the control field has been replaced by the advanced use of
the data field for supervisory communication. The next two octets form the protocol field,
which describes the contents of the data field. If the PPP frame is delivering a packet, the
protocol field contains the value 8021h. In PPP, this data field is called the information
field. It can be up to 1500 octets long, although a shorter value can be negotiated. Next
comes the frame check sequence field. As in Ethernet, this field is used to detect errors. If
the receiver detects an error, it simply discards the frame. There are no
acknowledgements.
a) Which fields form the PPP header?
Start flag, address field, control field, and protocol field form the PPP header
b) Which fields form the PPP trailer?
Frame check sequence field, stop flag form the PPP trailer
c) Figure 2-15 shows the final frame when an HTTP application transmits an application
message over an Ethernet switched network. Give the final frame if PPP is used instead
of Ethernet. Just say PPP header and PPP trailer. Do not give details regarding the fields
that make up the PPP header and trailer.
PPP header, IP header, TCP header, HTTP message, PPP trailer.
d) Give the final frame if the application is SNMP, which requires UDP at the transport
layer. The protocol is still PPP at the data link layer. Again, just say PPP header and PPP
trailer. Do not give details regarding the fields that make up the PPP header and trailer.
PPP header, IP header, UDP header, SNMP message, PPP trailer
e) When TCP sends a pure acknowledgement, it transmits a TCP message that only has a
header. The application layer is not involved at all in the acknowledgement process. Give
the final frame when the packet travels over an Ethernet LAN.
Ethernet header, IP header, TCP header, Ethernet trailer
3. a) In Figure 2-19, how many switched networks, physical links, data links, and routes are
shown? b) In Figure 2-19, how many data link, internet, transport, and application
processes in total are involved in the transmission?
3 switched networks.
6 physical links.
3 data links. (one in each switched network)
1 route.
b) Figure 2-19, how many data link, internet, transport, and application processes in total
are involved in the transmission?
9 data link processes (one on each host and switch and two on each router).
4 internet layer processes (one each on the two hosts and the two routers).
2 transport processes (one each on the source and destination hosts).
2 application processes (one each on the source and destination hosts).
4. Ethernet stations need Ethernet addresses. Do Ethernet switches need to have Ethernet
addresses too when they forward frames? Explain your reasoning.
No. For frame forwarding, Ethernet switches merely look at frame Ethernet
destination addresses and pass on the frames. This does not require switches to
have Ethernet addresses themselves.
<However, for managed switches, there has to be a way for switches to talk to one
another. This does require each to have an Ethernet address.>
Even Harder Thought Questions
1. Normally, only the transport layer standard is reliable. However, in Chapter 5, we will
see that 802.11 wireless LAN standards at the data link layer are reliable. Why do you
think this is so? (Hint: Review the logic in Figure 2-13 for clues.)