The University of Melbourne, School of Computing and Information Systems
COMP90007, Internet Technologies, Semester 1, 2021
Final Exam
• This examination is worth of 60% of your final mark. It is a 3 hour exam.
• There are 16 questions in the exam and a total of 6 pages including this cover page.
• You may access other materials during the exam. Advise on accessing materials for the exam: You only need a text editor to answer this exam’s questions in our view. No other materials are really needed. Calculators should not be needed as questions require simple calculations to answer. Referring to the book or other resources throughout this exam should not be of much help for this exam if you understood the contents of the subject and studied properly for the exam, and rather you may see that if you spend a lot of time looking through the pages of our book, you may be losing time that you could have used answering questions properly. This is a short answer exam and all answers should be in your own words, calculations, and your own drawings. You should not put citations to other documents/papers/books/material, nor copy/paste any material into the exam paper.
• Attempt all questions as partial marks will be available. No question requires writing lengthy answers. Please be clear and brief as you may lose points for unclear or redundant descriptions.
• The values in square brackets after questions show the marks allocated to each question.
• You are welcome to use the text editor of your choice to edit your answers. You do not need to repeat the questions themselves in your answers. Just make sure you use the right question number per answer. Type your answers: No handwritten answers! Drawings should also be done electronically and no question requires extensive drawing effort.
• Start your answer document with writing your student ID on top of your document, e.g., “Student ID: …”. Then write “COMP90007 Final Exam Answers” in the next line. Then go to the line after that and start with your answers. Start each answer with a proper heading such as “Question 1 Answer: …” and so on.
• Please answer questions in the given order and separate the answers with a few blank lines to give some space between the answers to different questions.
• Make sure to save your progress locally and regularly during the exam and at the end as a PDF version as well. Upload only that PDF version when you are finished. We recommend not leaving the uploading of the PDF version to the last minute!
Page 1 of 6
Question 1: [3 Marks]
We have discussed having acknowledgments in the Data Link Layer between two machines communicating to each other through a direct connection. Then we have discussed having acknowledgments with the TCP protocol again at Transport Layer. If we consider acknowledgements from the point of view of having reliability, among other reasons, please explain why we need these two different layers dealing with reliability over the Internet again and again? In a few sentences focusing on these two layers only mention what would be the reason for this situation above specifically.
Question 2: [3 Marks]
We may decide to connect two machines with a twisted wire which experts told us that it has the capability of having a 12KHz channel for communications. We are also told that we are given a new shielding with this wire, which makes it possible to communicate without any noise. Given these, what is the maximum data rate possible? Please briefly explain your answer.
Question 3: [4 Marks]
If we are asked to use Go-Back-N protocol at the Data Link Layer, please calculate the link efficiency of this protocol in terms of Link Utilization we saw in class under the following conditions. If N is set to 1, and frame size is 5000 bits, and propagation delay 500ms, and with a bandwidth of 5000bps. You can assume that time to transmit an acknowledgement is 0 seconds.
Question 4: [6 Marks]
We want to use the Cyclic Redundancy Check method for some data transmission as a means to deal with errors. Please compute the final data to be sent, given the following information. Show your steps and calculations. Original Data is 10110101 and your G(x) is x4 + x2 + x + 1. [Important Note: When showing your calculations you can use the following notation, but also note that the numbers given in the notation example below have nothing to do with the question above and the operations there should not be taken as an indication how your solution would be done. This example below is there simply to give you an example means that you can use to easily enter your solution in electronic form with your editor, while avoiding complicated drawings/diagrams.]
Sample Notation:
1001\10000111100 1001
————————- 0 0 0 1…
operation continues
Page 2 of 6
Question 5: [3 Marks]
For the following CSMA based MAC protocol flow chart we saw in class that is used to describe the process from the view of a machine that wants to communicate: what do we gain by having added the red “wait” stage to the protocol? Please explain in a few sentences.
Question 6: [4 Marks]
Given the Adaptive Tree Walk protocol we have seen in class, what happens with the following organization of stations and only stations A, C, F, and G want to send something. Please give steps in time, one per line, and explain briefly what happens in each time step with a sentence.
Question 7: [3 Marks]
Given the following network of routers each labeled with a character, e.g., a, b, etc, and assuming each connection, which is shown with an arc, has cost of 1, please draw the sink tree for the black sink node that is given in the middle. (Note: there could be more than one valid sink tree for any node given in a network. Also, you do not need to show your derivation but only the final sink tree.)
e Page 3 of 6
Question 8: [5 Marks]
Given the following network example for hierarchical routing that we saw in class (with the left subfigure below with regions and routers in each region) and the routing tables for 1A (i.e., two following subfigures on the right): Please find the new hierarchical as well as the full routing tables for router 1A, if the connectivity of the network would change in a way that connection from 1C to 3B is broken as well as the connection from 2D to 5C, but a new connection from 1C to 3A is established instead.
Question 9: [3 Marks]
What is the main benefit of aggregation in IP addressing that we saw in class, please briefly explain.
Question 10: [3 Marks]
If we already have IP addresses at the Network layer, which we can use to route any message to a particular machine from another one over the Internet, what is the benefit of having port numbers at the Transport layer on top of this addressing scheme. Please briefly explain.
Question 11: [3 Marks]
Briefly explain the difference between flow control and congestion control in networking.
Page 4 of 6
Question 12: [4 Marks]
With the following figure we have studied the concept of congestion in class. Please briefly explain why we cannot achieve the ideal response and in fact without proper measures we see a collapse in meaningful traffic on the network, using the following figure and information you have learned about the Internet and its protocols in class.
Question 13: [3 Marks]
Recall the Leaky Bucket algorithm we saw. If a sender has a burst data rate of 4000 bits per second for 15 seconds as data to send, and we have a bucket size of 25000 bits with an output rate of 5000 bits per second: Does the Leaky Bucket algorithm sustain its aim of regulating output or does it overflow and become inadequate and that we need a larger bucket, if so what should be its size? Briefly explain and show your calculations.
Question 14: [4 Marks]
Given the following Diffie-Hellman based protocol for creating a shared secret together, please name one attack type that we can use to undo this protocol and briefly explain in a few sentences how that attack would evolve step by step. Clearly label what the shared secret is and what each parameter you mention in your answer is used for and who is sending what message to whom. You do not need to draw figures to explain your answer.
Page 5 of 6
Question 15: [5 Marks]
One can use Cipher Block Chaining mode that we saw in class to make sure that the same input text does not lead to the same ciphertext all the time; the basic mechanism is depicted with a figure below where the transmissions continue presumably forever. Lets assume that in the middle of the following sequence a transmission error occurs with sending the second ciphertext and it arrived with an error to the recipient end and cannot be easily corrected at the receiver. What is the implication of this? Is there a solution for this problem at the ciphertext generation side? If there is one, please explain briefly. You do not need to draw a figure for this question and textual explanation is enough.
Question 16: [4 Marks]
I can create a simple secure communication method as follows: i) create a random bit string as a key for a particular transmission using my computer and throw away this key once I send this transmission ii) convert my plaintext transmission data that I want to send, by using its ASCII representation and converting it to a bit string as well iii) then compute the XOR of these two strings bit by bit (assuming they have the same length, i.e., for the two strings for the sake of simplicity for this question) iv) send the result string across to the other host who can undo the XOR operation. Is there a way to break this method, e.g., an intruder understanding this message by looking at this message somehow? If yes, give a way to break it, if not state the key reason why it cannot be broken. Then, briefly discuss the main strengths and weaknesses of this method.
…END OF EXAM
Page 6 of 6