Computer Networks

1. [IP header] (5 marks)
A router R1 acts as a gateway to a VPN tunnel. At the other end of the tunnel is
another gateway R2. R1 and R2 are not directly connected to each other. Suppose that
the Router R1 receives an IP datagram that needs to traverse the VPN tunnel. This
implies that this incoming datagram needs to be encapsulated in another IP datagram
before it can be sent over the VPN tunnel. However, R1’s path MTU information
indicates that this incoming datagram requires fragmentation. Now, router R1 has two
options to deal with this situation:
Option 1: Encapsulate the original datagram and then subject the encapsulated packet
to fragmentation.
Option 2: Fragment the original datagram and then encapsulate each resulting
fragment.
Which of the two options should the router R1 opt for and why? (200 words)
2. [Intra-Domain Routing]
a) (5 marks) So far in this lecture series, for intra domain routing scenarios, we have
assumed that the shortest path is the least cost path where cost is may be defined
in terms of number of hops, delay etc. This scheme simply builds the overall cost
of a path by adding individual link costs along the path until all links for the path
are considered. There are many networking environments in which this strategy of
selecting the best path using least cost is not applicable. Identify and explain one
scenario where this routing strategy may not work. (200 words)
b) (4 marks) An organization has a network that uses older protocols such as RIP-1.
What reasons might motivate this organization to upgrade to newer protocols that
use VLSM and CIDR? What issues should be considered in deciding whether to
upgrade to newer protocols that support VLSM and CIDR? (150 words)
c) (10 marks) Network engineers examine many features when deciding which
routing protocol to use. Investigate and briefly describe 5 features that characterise
routing protocols which will help a network engineer distinguish one protocol
from another. (500 words)
3. [CIDR]
a) (2 marks) Summarize the following range of IP addresses by ensuring the
summaries do not include any unnecessary address ranges.
172.47.30.0/24
172.47.31.0/24
172.47.32.0/24
172.47.33.0/24
b) (3 marks) In some situations, a subset of the more specific routes needs to be
advertised in addition to the aggregate? Explain why with the help of an example.
(Word limit: 150 words)
c) (6 marks) Consider the network configuration shown below. The customer
network is allocated the address space 140.2.0.0/16.The Border Router A of the
customer network advertises an aggregate route for that address range and
internally uses subnetting on the entire third octet space to create 255 subnetworks
addresses, not all of which are actually allocated and assigned since the
customer currently does not need so many networks. The ISP router B advertises
5
the default route (i.e., 0.0.0.0/0) – an aggregate route as well – back to the
customer network. This implies that all traffic destined to target prefixes not
located in the customer’s network must be sent via the ISP’s network. Identify and
explain any one problem this configuration could lead to and provide a remedial
solution. (200 words)
{Customer networks: 140.2.1.0/24, 140.2.2.0/24, 140.2.5.0/24}
d) (5 marks) Consider the scenario depicted in Figure below. For network design to
be resilient, access networks are frequently connected to the core through multiple
border routers. Routers R1 and R2 are border access routers responsible for
aggregating routes. The configuration is such that routers R1and R2 both advertise
an aggregate route for 12.2.0.0/16 with the same metric values. Router R1
receives the aggregate route advertised by router R2 via the core network. R1 now
has two routes to 12.2.0.0/16: its own local route and the route received from R2.
Assume that R1 prefers R2’s aggregate route over its local route. As a result,
router R1 stops advertising the aggregate route and instead, re-advertises the route
from R2 after incrementing its metric. Describe what problem you found in this
configuration. Explain how this problem can be fixed. (200 words)
e) (3 marks) An ISP is granted a CIDR block of addresses starting with
120.60.4.0/22. The ISP wants to distribute these blocks to 100 organizations with
each organization receiving 8 addresses only. Design the sub-blocks and give the
6
slash notation for each sub-block. Find out how many addresses are still available
after these allocations.
4. [BGP]
a) (5 marks) BGP uses the services of a reliable transport protocol (TCP) to
transport its messages. What changes do you foresee in BGP design if it were
made to operate over an unreliable connectionless transport protocol such as the
UDP? (300 words)
b) (7 marks) Consider the routing scenario in the Figure shown below. An
autonomous system (AS1) is connected to the Internet via two connections. The
router R1 in Sydney is running EBGP with one provider over a T1 circuit, and the
router R2 in Melbourne is running EBGP with another provider over a T3 circuit.
Inside AS1, the two border routers R1 and R2 are running IBGP and EBGP, but
are not physically connected. All communication between R1 and R2 go via
internal routers R3, R4 and R5. The internal routers R3, R4, and R5 are non-BGP
routers and exchange routes via an IGP with all other routers in the AS. In the
current scenario, R1 and R2 are both receiving full routes from their parent
providers. These routers are also injecting a 0/0 default route inside AS1. Assume
also that AS1 runs the primary/backup technique to enable the MEL T3 link to be
the primary. The SYD link will be used as a backup, and therefore all outbound
traffic that reaches R1 will now need to be directed back towards R2. Interior
routers R3, R4, and R5 do not see any of the exterior routes and follow defaults
towards R1 and R2 according to the IGP metric. Describe what problem can occur
in this scenario. Devise some alternate methods to avoid this problem. (300
words)
7
5. (5 marks)
A company has registered to receive 8 different channels of a high bandwidth video
stream from a central facility via 8 different multicast groups. The company’s
network has the capability to support three active streams concurrently. The company
has support for multicasting and uses version 1 of IGMP to manage its active groups.
What problem can stem from the use of IGMP version 1 in this scenario? Explain.
(250 words)

Use the order calculator below and get started! Contact our live support team for any assistance or inquiry.

[order_calculator]