Final Spring2015
Final Spring2015
Final Spring2015
FINAL
May 21, 2015
150 minutes
Name:
Student No:
Show all your work very clearly. Partial credits will only be given if you carefully state
your answer with a reasonable justification.
Q1
Q2
Q3
TOT
1)
a) (6 pts) Suppose you are developing a multimedia networking application which requires a
minimum data transfer rate from the network in order to function effectively. Which of the
two transport layer protocols, TCP or UDP, will you prefer? Justify your answer.
b) (5 pts) Complete the following table.
Router Switch Hub
Layer of operation
(1,2,3,4,5)
Isolates network
traffic (Yes,No)
Does not require
configuration from
network
administrator
(Yes,No)
May implement
OSPF (Yes,No)
Has store and
forward capability
(Yes,No)
In the following table indicate to which subnets the following hosts with given IP addresses
belong. If there are some hosts that do not belong to any of the subnets, mark them with a “-
``.
Host IP Address Subnet
A 139.179.12.118
B 139.183.12.16
C 140.183.12.115
D 140.183.190.55
E 140.183.224.88
b) (6 pts) The largest IP router can hold 200,000 entries in its forwarding table while the largest
Ethernet switch can hold 1,000,000 entries in its forwarding table. However, a large IP
network can support many more hosts than a large Ethernet network. Why?
c) (12 pts) The Dijkstra algorithm that we studied in class is given below. Dijkstra’s algorithm
computes all least-cost (i.e., shortest) paths from node u to all other nodes in the network
where the cost of a path (x1,x2,x3,…,xp), denoted as pc(x1,x2,x3,…,xp), is given by the sum of
all the link costs along the path; that is,
pc(x1,x2,x3,…,xp) = c(x1,x2) + c(x2,x3) + … + c(xp-1,xp).
Now assume that we modified the definition of the path cost as the product of the link costs
along the path; that is,
pc(x1,x2,x3,…,xp) = c(x1,x2) * c(x2,x3) * … * c(xp-1,xp),
where the costs are real numbers in the interval (0,1). Considering that the costs are
probabilities of links being intact (i.e., working), that the costs of links are independent of
each other, and the task is to find the most reliable (i.e., highest cost) paths from node u to
all other nodes in the network, use the following table to give a Modified Dijkstra Algorithm
which solves this new problem.
Hint: Note that -log(a * b) = -(log a + log b).
3 4
1
5
3 4
Give the evolution of the distance tables with respect to destination 5. Specifically, give the
distance table entries for destination 5 at nodes 1-5, for t = 0.1, 0.5, 1.1, 2.1, ..., until all
distance vectors stabilize. Present your final answer in the table given below where D i (j) is
the distance vector element denoting the distance from i to j.
b) Suppose that nodes A, B and C are connected to the same Ethernet. Assume that each of
these three nodes is trying to retransmit a frame. These frames have already experienced 2,
3, and 4 collisions, respectively, i.e., collision counters are 2, 3 and 4 for nodes A, B and C.
Assume further that all other nodes on the Ethernet are inactive.
i. (5 pts) Assume that all three nodes detected the last collision simultaneously. Node C
chooses 5 as the random number according to the exponential backoff algorithm.
What is the probability that A will be the first node initiating a retransmission
attempt?
ii. (5 pts) Consider part i. if node C chooses 2 as the random number (instead of 5).
What is the probability that B will be the first node initiating a retransmission
attempt?
c) (6 pts) Consider a 100 Mbps 100BaseT Ethernet network with a star topology composed of
5 nodes where each node has a point-to-point connection with a hub in the middle. The
distances from the 5 nodes to the hub are given by 100m, 200m, 500m, 900m and 1000 m,
respectively. Calculate the minimum frame size that this LAN can support so that the
CSMA/CD protocol will function correctly. Assume that the speed of propagation is 2x108
m/s.
d) Consider the network given in the figure below, where hosts X, Y, Z, W are connected
conn via
Switches 1, 2 and 3. Initially assume that the switch ((i.e., bridge) tables
les of all switches are
empty. Recall that each switch applies the self learning algorithm to learn which hosthosts can
be reached through
rough which interfaces. Answer the questions below when the following events
happen sequentially over time.
i. (3 pts) First, X sends a packet to Z. Which switches learn where X is, and does Y's
network interface see this packet?
ii. (3 pts) Z now sends a packet to X (after X hass sent to Z). Which switches learn
where Z is, and does Y's network interface see this packet?
iii. (2 pts) Next, Y sends a packet to X (after the above two packets). Which
W switches
learn where Y is, and does Z's network interface see this packet?
iv. (2 pts) Finally, Z sends a packet to Y (after the above three packets). Which
W switches
learn where Z is, and does W's network interface see this packet?