Computer Science Questions

Department of Computer Science

Q1                                                                              

In a link state routing algorithm, link state packets are sent to all routers (3 marks) in the network when the state of a link changes. How is this accomplished  reliably?  

How and why are link state packets sequence numbered?     (3 marks)

Application level protocols are those used directly by user applications to (5 marks) achieve their goals. Examples include FTP – the file transfer protocol, HTTP – the hypertext transfer protocol, SMTP – the simple mail transfer protocol. Using FTP and HTTP as examples, briefly outline how they communicate with their peer processes on remote hosts, giving attention to how commands and responses are communicated and how the actual data requested is transferred. You should state where they are similar and where they differ.

You have been tasked to select a suitable routing protocol for an enterprise (5 marks) which owns a very dynamic large network, discuss which routing protocol classes you think is better and why. Justify your answer based on the following metrics (protocol type, distance metric, route selection algorithm), stating the strengths and weaknesses.                                           The figure below shows an inconsistent data rate flow (bursty traffic) that (4 marks) a source transmits in the following data pattern: continuous 1024 Mbps data for 20 seconds,  then a gap (an empty time slot) for 15 seconds (no data traffic during this time slot),then, continues 224 Mbps data for 15 seconds.

What is the total amount of data that has been transmitted in a 50 second time span?Explain how the leaky bucket algorithm maintains a fixed data rate flow, and what is the fixed data rate in 50 sec? Justify your answer by drawing a new data rate flow graph.                                              

Q2                                                            

The design of many Internet protocols is driven by the need to reduce (3 marks) delay and increase throughput. Explain how the following types of delay  can be reduced (if at all).transmission delay iii. queueing delay                          Error detection is an essential component of the Internet. Compare and (4 marks) contrast error detection as provided by link layer protocols with that provided by IP and TCP.        Why has IP version 6 dropped error detection capabilities that were (3 marks) present in IP version 4?

Cyclic Redundancy Check (CRC) error detection method treats packets (5 marks) of data to be transmitted as a polynomial. The following is the generator and data that the source node wants to send. Generator G(x) : X5 + X2 

What is the CRC that the source needs to append to the data?  

What is the complete frame (data + CRC) that the sender is required to send?  

Show your working for both i and ii.

Transport layer protocols are concerned with end-to-end delivery of data. (5 marks)

                       Transmission Control Protocol (TCP) is the main connection-oriented          

transport protocol used in the Internet.

Why does TCP not choose a fixed window size for its connections? How does it signal changes in the window size?

Why is a fixed timeout value not a good choice and how does TCP set its timeout values?

Q3                                                                              

A software company hires you to build an email client software that can (6 marks) read and send emails. Discuss which network protocols your software  will need to be able to communicate in and why.          Your company plans to use the 10.X.Y.0/24. Choose the X to be your day (8 marks) of birth and Y to be your month of birth. For example, if you were born on the 22nd of October, then your network should be 10.22.10.0/24. You are tasked with subnetting that network into four sub-networks. The first sub-network should be able to accommodate 64 hosts. The second needs to be able to hold 61 hosts. The remaining networks are to be of equal size and to be able to accommodate 17 hosts each. For each subnet, provide its subnet mask, network identifier, broadcast address and range of usable IP addresses.            Suppose that you have a machine that is configured to use a local DNS (6 marks)

server “A” which is configured to use the DNS server “B” provided by your ISP. “B” is configured to query root DNS servers. 

Provide the operations that will result from your query to get the IP address of www.an_example_website.com, assuming that the caches of all machines in this scenario have been wiped out.  

Assuming that the caches have been populated from the query in (i), provide the operations that will result from your query to get the IP address of users.an_example_website.com

Q4                                                            

Consider the following topology of routers where distance metric (8 marks)

values between directly connected routers are indicated on the links:

Distance vector routing is being used. All routers have just been booted. Router R1 receives the following updates (note that optimal routes have not been identified yet):

Advertisement from R2 is R3:2, R4:1.

Advertisement from R3 is R2:2, R4:3, R5:4, R6:1.

Provide the distance vector that R1 will advertise after processing these  updates and explain how R1 calculates it.            

What is the distance vector that R1 will advertise once the network has       (6 marks) converged?

With “split horizon” in place, and assuming that the network has already (6 marks) converged, what is the distance vector of R6? What is the distance vector that R6 will send to R3 and to R7?                                      

                     END OF PAPER