Comp6461 – Fall 2014 Theory Assignment 3 - Page 1 COMP 6461 - Fall 2014 Theory Assignment 3 1) What is the main difference between Pure Aloha and Slotted Aloha? Are there any circumstances where Pure Aloha would perform better than Slotted Aloha? If so, give such circumstances/conditions. If no, explain why Pure Aloha could never perform better that Slotted Aloha. 2) Given a network with 9 routers as shown in the figure below. Assume the utilization of center-based spanning tree construction, where router D is assigned as the center (or root) router. Assume routers join the tree in the following order: C, A, H, B, F, G, E, and I. Show the final constructed spanning tree for that given network. 3) Consider two substitution ciphers. One adds a value of i to the ASCII code of the plaintext character. The other adds a value of j to the plaintext character. All additions are modulo 256. Now consider a double-encryption method that adds i to each plaintext character and then adds j to the resulting ciphertext character to get another ciphertext character. Again, all calculations are modulo 256 How much more secure is this double encryption when compared with either single-encryption method? Explain your answer. 4) Consider the bit string 001011010101000011111101001101 and the key 10110. Use the key to encrypt and then decrypt the string using bit level ciphering. 5) Suppose you were trying to crack an encryption method that used a 64-bit key. Assuming a brute force attack, how many keys per second must you try to crack the code in 30 days? A C D G B E F H I Comp6461 – Fall 2014 Theory Assignment 3 - Page 2 6) Three broad classes to multiple access techniques exist, which are: channel partitioning, random access and taking turns. In general, it is assumed that any network would use one of these techniques or the other. However, cable Internet access networks utilize all three techniques. Describe how this is done by such networks, and explain how the utilization of any of these techniques does not conflict with the utilization of the other two. 7) In general, fully-connected topology is exhaustive and very unlikely to be used for the construction of a network. a. Are there any clear and significant advantages of such topology? Explain clearly. b. Explain why it is unlikely that such topology be used for network construction. You should clearly indicate the major disadvantages of such topology. c. While this topology seems to make little sense for the construction of networks, in the general terms, it is actually used as part of Data Centers to connect Tier-1 and Tier-2 switches. Explain clearly the major advantages of such utilization in data centers. You should provide some example that shows, through some numerical values, the obvious advantages of such utilization. 8) Assume the utilization of Hamming Codes for single-bit.

1. Comp6461 – Fall 2014 Theory Assignment 3 - Page 1
COMP 6461 - Fall 2014
Theory Assignment 3
1) What is the main difference between Pure Aloha and Slotted
Aloha? Are there any
circumstances where Pure Aloha would perform better than
Slotted Aloha? If so, give
such circumstances/conditions. If no, explain why Pure Aloha
could never perform better
that Slotted Aloha.
2) Given a network with 9 routers as shown in the figure below.
Assume the utilization of
center-based spanning tree construction, where router D is
assigned as the center (or
root) router. Assume routers join the tree in the following order:
C, A, H, B, F, G, E, and
I. Show the final constructed spanning tree for that given
network.

2. 3) Consider two substitution ciphers. One adds a value of i to
the ASCII code of the
plaintext character. The other adds a value of j to the plaintext
character. All additions are
modulo 256. Now consider a double-encryption method that
adds i to each plaintext
character and then adds j to the resulting ciphertext character to
get another ciphertext
character. Again, all calculations are modulo 256 How much
more secure is this double
encryption when compared with either single-encryption
method? Explain your answer.
4) Consider the bit string 001011010101000011111101001101
and the key 10110. Use the
key to encrypt and then decrypt the string using bit level

3. ciphering.
5) Suppose you were trying to crack an encryption method that
used a 64-bit key. Assuming
a brute force attack, how many keys per second must you try to
crack the code in 30
days?
A
C
D
G
B
E
F
H
I
Comp6461 – Fall 2014 Theory Assignment 3 - Page 2
6) Three broad classes to multiple access techniques exist,
which are: channel partitioning,

4. random access and taking turns. In general, it is assumed that
any network would use one
of these techniques or the other. However, cable Internet access
networks utilize all three
techniques. Describe how this is done by such networks, and
explain how the utilization
of any of these techniques does not conflict with the utilization
of the other two.
7) In general, fully-connected topology is exhaustive and very
unlikely to be used for the
construction of a network.
a. Are there any clear and significant advantages of such
topology? Explain clearly.
b. Explain why it is unlikely that such topology be used for
network construction.
You should clearly indicate the major disadvantages of such
topology.
c. While this topology seems to make little sense for the
construction of networks, in
the general terms, it is actually used as part of Data Centers to
connect Tier-1 and
Tier-2 switches. Explain clearly the major advantages of such
utilization in data
centers. You should provide some example that shows, through
some numerical

5. values, the obvious advantages of such utilization.
8) Assume the utilization of Hamming Codes for single-bit error
correction.
a. What is the total number of bits that need to be transmitted if
the original data
string has 11 bits? Which positions will be covered by the
different parity bits?
b. What is the total number of bits that need to be transmitted if
the original data
string has 19 bits? Which positions will be covered by the
different parity bits?
9) Token Ring LANs have the clear advantage of avoiding
collisions, which is surely not
the case for networks, i.e. Ethernet, that use buses as the main
segment to connect the
different devices. In spite of that, Ethernet LANs are capable
today of providing a
superior performance in comparison to Token Ring LANs.
Explain how the Ethernet was
able to achieve that regardless of the collision issues. In
particular, your answer should
consider: 1) the disadvantages of Token Rings, and 2) Switched

6. Ethernet.
Comp6461 – Fall 2014 Theory Assignment 2 - Page 1
COMP 6461 - Fall 2014
Theory Assignment 2
1) E-mail requires both sender and receiver mail servers to
communicate directly. Further,
all communications must be made between these two servers
using SMTP.
i) Is it possible to allow intermediate servers as part of this
communications? If so,
what are the main advantages of doing so? If no, why is that
infeasible, or what
are the disadvantages of utilizing it if it was feasible?
ii) Considering webmail, are there any cases when SMTP
between the two mail
servers can be replaced by HTTP? Explain why, or why not.
2) Client-Server architecture can be used for file transfer;

7. however it is assumed/said to be
inferior in comparison to P2P for such operations. Considering
a small number of
interacting hosts that need to share files, is it true that client-
server would perform badly?
Explain clearly your answer. You must indicate why the number
of interacting
hosts/peers is significant in determining whether or not client-
server is suitable for file
transfer.
3) Show, through a detailed scenario/example, why GBN
protocol would fail in case the
sender’s window size exceeds 2
k
-1, where k is the number of frame bits for the sequence
number.
Considering that the window size is set to what it is expected to
be. Is there any
possibility that GBN could produce the same performance
obtained by the Unrestricted
Protocol? If no, explain clearly why this is infeasible. If so,
explain clearly how, or under
what conditions, GBN could produce the same performance of

8. the Unrestricted Protocol.
4) Explain how UDP frames can end up TCP frames being
delayed, theoretically,
indefinitely! If so, what would you propose as changes to UDP
to mitigate this problem?
Your solution must mainly keep the advantages/purpose of
UDP, while mitigating the
problem at hand.
5) Show, through an example, how checksum could be
inconclusive of error detection (i.e.
does not guarantee that errors can be detected). In your
example, assume transmitted data
is broken into 24-bit chunks by the protocol utilizing checksum.
In case errors are
detected by checksum, does that fully (100%) guarantee that
errors must have actually
occurred?
Comp6461 – Fall 2014 Theory Assignment 2 - Page 2

9. 6) Show through an example (sketch the scenario), how TCP
uses the credit system to
ensure that a sender would not overflow a receiver. In your
scenario, assume that initial
communication establishment resulted in sender (host A in
general since it also behave as
a receiver) starting with frame # 64, while the receiver (host B
in general since it also
behaves as sender) starts at frame # 192. Assume that the data
size of all these exchanged
frames is 100 bytes. Further, assume that at start, each of the
hosts allows the other one a
credit of 200. The rest of the scenario is up to you, however,
you need to show at least
one case when both of them coincidently set the credit to 0 at
the same time.
7) Some may strongly argue that it is much better to use
network-assisted congestion control
instead of end-to-end congestion control. Do you support such
argument? Clearly explain
why or why not, and indicate the advantages of your choice over
the other one.
Why does TCP use end-to-end congestion control over the
Internet?

10. 8) With VC networks, the frame does not include the receiver
address. Explain clearly how
this would still work! Give a detailed scenario that explains
how deliveries to the receiver
can still be obtained. Show all needed tables at the routers in
support to your scenario.
9) What is CIDR? What are the main advantages of using
CIDR? Are there any major
disadvantages to this technique?
10) With NAT routers, many machines over the Internet would
have the same IP address!
Explain how this would still work. Explain also how a P2P non-
NAT host can establish a
connection with another peer located under a NAT.
Comp6461 – Fall 2014 Theory Assignment 1 - Page 1
COMP 6461 - Fall 2014

11. Theory Assignment 1
1) Circuit Switching aims at providing a better service through
the reservation of the circuit
(i.e. circuit is dedicated). Now, considering only the
perspective of the communicating
users over a Circuit Switching network (i.e. you should not be
concerned with the entire
utilization of the network or the advantages to other users), is it
possible that Circuit
Switching may actually end up harming its users instead of
providing a better service to
them. If yes, provide a scenario/case that shows that. If no,
explain why this service will
indeed provide the best service to its users at all times.
2) With DSL ISPs, dedicated lines usually connect to the local
offices to the location where
the service is provided (i.e. residential homes). In contrast, with
Cable ISPs, the
connection is shared between multiple homes in a
neighborhood. In spite of this
configuration that is clearly to the benefit of DSL users, service

12. provided by Cable ISPs
may still be superior to the one provided by DSL providers.
Explain the reasons behind
this. Further, if you are hired by a DSL provider, and taking
into account that changes to
the company’s infrastructure (i.e. wiring) is quite costly, what
would you propose in
order to speed up the provided service while balancing the cost.
3) HTP provides two connections methods: persistent and non-
persistent. While there are
clear advantages of persistent-HTTP, it is rather unclear
whether non-persistent can be of
any use. Conduct a small research to find out why these two
modes of connections are
provided. In specific, you should find out some of the
advantages and disadvantages of
each of them, hence justifying their concurrent existence.
4) A successful attack to the Internet DNS would be
devastating. Explain what type of
attacks can be made towards DNS. Why, to-date, such attacks in
practice have not been
successful? In your answer, you should consider caching in

13. particular. Why such
technique has not only proven to provide better performance,
which is its original goal,
but also protection against security attacks.
5) Perform a Traceroute between your machine and any other
host/server on the Internet,
preferably an overseas server. Provide snapshots of what was
returned and analyze the
returned information. In specific, you should comment on any
behavior that looks either
different or unusual. Indicate the number of routers between
your machine and the
targeted host/server.
Comp6461 – Fall 2014 Theory Assignment 1 - Page 2
6) Suppose two hosts, A and B, are 10,000KM apart and are
connected by a direct link of
rate R = 5 Mbps. Assume further that the propagation speed
over the link is 2.5 x 10
8
, and

14. that the packets to be transferred are of size 2 Mbits.
a. What is the propagation delay to send one packet from A to
B?
b. What is the transmission delay to send 8 packets from A to
B?
7) Now assume some modifications to the network specified in
Question 6 above, where 2
routers are installed between A and B at more or less equal
distances. Further assume the
link from A to the first router has been replaced by another link
of rate R=4 Mbps, the
link between the two routers was kept as before, and the link
between the second router
and B was replaced by a link of rate R=6 Mbps. Assume
negligible processing and
queuing delays in the routers.
a. What is the propagation delay to send one packet from A to
B?
b. What is the total transmission delay to send 2 packets from A
to B?
c. How much time is needed to send one file of size 20 Mbits
from A to B?

15. 8) BitTorrent uses a trading scheme referred to as tit-for-tat.
While some researchers argued
particular deficiencies in that scheme, others argued that if it
was not for tit-for-tat, it is
likely that BitTorrent would not even exist today. Explain both
points of view. In other
words, you should indicate the disadvantages of the scheme, as
well as why it is still
crucial to have such a scheme, or an alternative one, for
BitTorrent to exist.
9) (Textbook Question P.22 – Page 177). Consider distributing a
file F of size 15 Gbits to N
peers. The server has an upload rate of us= 30 Mbps, and each
peer has a download rate
of di = 2 Mbps and an upload rate of u. For N = 10, 100, and
1000 and u = 300 Kbps, 700
Kbps, and 2 Mbps, fill the following table by the minimum
distribution time for each of
the combinations of N and u for both client-server distribution
and P2P distribution.
Client Server
N

16. 10 100 1000
u
300 Kbps
700 Kbps
2 Mbps
Peer to Peer
N
10 100 1000
u
300 Kbps
700 Kbps
2 Mbps