Practice exercises on wireless and mobile networks

Problem 1.
In an IEEE 802.11b LAN, RTS frames are 20 bytes long. CTS frames are 14 bytes long. ACK frames are 14
bytes long. The transmission rate is 11 Mbps. Suppose the 802.11b stations are configured to always reserve the
channel with the RTS/CTS sequence before sending data. Suppose station A wants to transmit 1000 bytes of user
data, and all other stations are idle at this time. Ignore propagation delay, and assume that there were no bit errors.
See http://www.ece.virginia.edu/mv/edu/4457/lectures/Chapter6/RTS-CTS.jpg.
a. Express the total time incurred for a successful transmission of the 1000-byte user data from station A
to the access point (express this time as a function of SIFS and DIFS). [Hint: Include the interframe
spacings used in the RTS/CTS exchange; “successful” implies ACK reception; remember to include
the frame header/trailer.]

b. If the total time to send 1000 bytes of user data is 867 µs , what is the efficiency with which the 11
Mbps wireless link is being utilized? Express as a percentage.

Problem 2.
Assume the minimum and maximum contention window sizes in an 802.11 LAN are 31 and 1023 slot times,
where slot time (denoted SlotTime ) is 20 µs . SIFS is 10 µs , and DIFS is SIFS + 2 × SlotTime . Consider how
your answer to Problem 1(a) will change if the medium is busy when the sending station (station A) first sensed the
medium. Specifically, if the station is able to transmit its frame on its first attempt without collisions, what is the
average additional time required for successful transmission of the 1000 byte payload relative to your answer in
Problem 1(a)?

Page 1 of 4
Problem 3. [Synthesis problem]

Hosts HA1 through HA10

LAN A 802.11
wireless 802.11
LAN B host Access
Ethernet Point
2 (AP1)
Bridge 3
1 Bridge 1 1 4 LAN E
2
LAN D 1 2
3 Bridge 4
2
1 3 LAN C
Bridge 2 IP router HC1
1
2 Internet

Above is an extended version of an Ethernet switched/hubbed network that we used in Assignment 8. Recall that
all networks marked as LANs are hubbed Ethernet LANs, and all bridges are Ethernet switches. Hubbed LANs have
multiple hosts (e.g. HA1 ... HA10 on LAN A) and one or more hubs. Assume that HA1 through HA10 are the MAC
addresses of the host interfaces. The naming convention shown for the MAC addresses on LAN A is used for host
interfaces on the other LANs, e.g., if there are 20 hosts attached to LAN E, their host interface MAC addresses are
HE1 through HE20. Hosts are referred to by their MAC addresses.
An IEEE 802.11 access point has been connected to port 4 of bridge 4 via an Ethernet link. An IP router has been
connected to port 3 of bridge 2. Assume switch (forwarding) tables are all empty at the start, and bridges slowly learn
MAC addresses and add entries to their switch tables.
a. If host HA1 sends an Ethernet frame destined to host HC1, the frame will reach all LANs because no
bridge will have an entry corresponding to HC1 in its switch table. Therefore the access point will
receive the Ethernet frame. Will it forward this frame on its wireless interface?

b. If the 802.11 wireless host shown in the network is associated with access point AP1 (i.e., it has sent an
Association Request and received an Association Response), and it sends an 802.11 frame carrying an
IP datagram destined to a host connected to the Internet as shown in the network above, then what val-
ues are carried in the Address1, Address2, and Address3 fields of the 802.11 header?

c. If the wireless host shown in the network is associated with access point AP1, and it sends an 802.11
frame carrying an IP datagram destined to host HA5, then what values are carried in the Address1,
Address2, and Address3 fields of the 802.11 header?

d. How does the wireless host decide whether to send an 802.11 frame carrying an IP datagram to the
router or directly to a host within its subnet?

Page 2 of 4
 e. When the access point receives the 802.11 frame sent by the wireless host in part (b), it sends the pay-
load of the frame in an Ethernet frame to bridge 4. What address value is placed in the source MAC
address field of this Ethernet frame?

f. The wireless host uses DHCP to acquire an IP address, subnet mask, and default gateway. It is assigned
the IP address of the router’s port 1 as the default gateway IP address. What can be said about the IP
address and subnet mask assigned to the wireless host (for its 802.11 interface) when compared to the IP
address and subnet mask of the router’s port 1?

g. Do the bridges in the network learn the MAC address of the wireless host and store entries correspond-
ing to this MAC address in their switch tables?

h. Assume a second access point (AP2) is connected to another port (say port 5) of bridge 4 (not shown in
the figure), and the wireless host moves to the coverage area of this access point (AP2). Assume that the
wireless host had sent frames in its original position (when it was in the coverage area of AP1) before it
moved into the coverage area of AP2. Will Ethernet frames destined for the wireless host be sent out by
bridge 4 on the wrong port (i.e., port 4), or will these Ethernet frames be properly forwarded to port 5?
Answer this question for two cases: the frame destined to the wireless host arrives at bridge 4 before and
after the wireless host has itself sent an Ethernet frame destined to some other host within the network
or on the Internet from its new location?

Problem 4. (
Experiment with the Java applet, 802.11 CSMA/CA WITH Hidden Terminals (Chapter 6), in the Textbook Web
site (http://wps.pearsoned.com/ecs_kurose_compnetw_6/216/55463/14198700.cw/index.html). First test the applet
out to gain an understanding of the applet. Here is the experiment that we will be repeating: First click on the start
button, and then in rapid succession, click on Station 1, Station 2, and Station 3.
a. Do you see collisions of the RTS frames? Now try the same experiment with the “CSMA/CA without
hidden terminals” applet. Do you see collisions of the RTS frames in the latter case?

b. Return to the With Hidden Terminals applet. After the initial set of collisions with the RTS frames,
when the transfer of a data frame starts, do you see any collisions with that data frame?

c. After an RTS was successful, what frame is sent and by which node?

Page 3 of 4
 d. When one station is sending a data frame, what status is shown for the other stations? What is NAV? In
what field of the IEEE 802.11 header was the information related to NAV communicated to the other
stations?

e. What frame follows the transmission of a data frame by a station?

f. Run the experiment specified in the problem description three times (with resets in between). In each
case, as soon as the RTS collisions end and a data frame is sent successfully, Pause the applet, and use
PrtSc to capture the screen. Save the file in a Word or similar document. Answer two questions about
each run: (i) how many RTS collisions occurred before success? (ii) what does the value in the Queue
indicate? Your answers should be supported with the information in the screen captures. Save your doc-
ument as a PDF file, and submit on Collab under Assignments - use the naming convention
TeamN_Lastname1_Lastname2_Lastname3_Lastname4.pdf.

Problem 5.
Let’s say the FCC allocated a total of 40 MHz for a new service and divided this into two 20 MHz bands, one for
the upstream direction and the second for the downstream direction. This service is bidirectional, with each commu-
nication session requiring one 20 kHz channel in each direction. It is an FDMA/FDD system. Assume that 20 chan-
nels are set aside in each direction for control messages (which means these channels are unavailable to carry user
data). Assume that a service provider was able to purchase the full 40 MHz band for a coverage area of 96 cells to
provide this service. The service provider chooses a reuse factor of 4. What is the maximum number of simultaneous
bidirectional communication sessions that can be supported by this service provider?

Problem 6. [unrelated parts]

a. Assume a cell phone user with a Manhattan phone number, e.g., 212-512-8501, is visiting Charlottes-
ville where the area code is 434. If a Washington DC phone customer with a 202 number calls 212-512-
8501, how is the call routed? Specifically which mobile switching centers are involved?

b. What number is used to route calls from the home network to the visiting network in cellular networks?

c. Do telephone network switches use aggregated (summarized) address ranges in their routing (switch)
tables? In other words, is there a single entry in the Washington telephone network switch for routing
calls to all telephone numbers with area code 212?

Page 4 of 4