COS 116 The Computational Universe Laboratory 10: Virus and Worm Propagation in Networks You learnt in lecture about computer viruses and worms. In this lab you will study virus propagation at a quantitative level. You will use a simple simulation model to gather data and then use Microsoft Excel to understand the data. The model simulates a network of interconnected computers reminiscent of the internet but with fewer computers (up to 100,000). When you set all relevant parameters and click “Run Simulation,” it simulates the spread of the virus/worm in the network and outputs some numbers for you to analyze. Obviously, there are close analogies to spread of diseases in humans, so this lab gives you some insight into that as well. Lab submission: Submit by May 10 in Room 410 of the CS building (Mrs. O’Leary’s office) during office hours. Turn in all the Excel charts that are followed by a star (*) in the lab. Make sure each chart is labeled with the Experiment number and Question number, as well as any relevant additional information (e.g. particular options). You will not receive credit for charts that are unlabeled. Do not turn in the charts without a star; they are just for you to visualize the simulation. Also, turn in answers to all the questions posed in the body of the lab and in the “Additional Questions” section. Introduction: Using the Simulator 1. Download this file to your Desktop: http://www.cs.princeton.edu/courses/archive/spring06/cos116/vsim.exe 2. Double-click the file to run the simulator. Ignore any security warnings. The simulator has a variety of options to adjust. These options are explained below. Note that there is an element of randomness in the simulator, especially when using the social network settings, so your results may not look exactly like your neighbor’s.Figure 1: Simulator interface Here is a brief explanation of all adjustable parameters. • Network size: The number of nodes in the network. This can be interpreted as the number of computers in a computer network or the number of people in a social network. • Network type: In a fully-connected network, every node is directly connected to every other node. In a social network the connections are generated according to a randomized scheme so that the resultant network resembles actual social networks that have been studied. Though the details of the social network are interesting (feel free to ask us questions) they are not relevant for this lab. You just need to know that each node is directly connected to at least the number specified in “Minimum Friends” though some nodes may be connected to much more than this minimum. (In social networks, these are the popular people.) • Initially vulnerable: The percentage of computers that are initially vulnerable to infection. This models the possibility that some computers may already be patched against a weakness, or that some people are immune to a disease. • Initially infected population: The number of computers that are initially infected. • Time to install: The time it takes for an infected computer to begin spreading the virus to other computers. • Time to spread: The time between infection attempts by an infected computer that is spreading the virus. • Time to repair: Time from infection until the virus is removed and the computer is immunized. How infection works: An infected computer spreads the virus by randomly choosing a computer to which it is directly connected every "Time to Spread" seconds. If that computer is vulnerable, then it will begin spreading the virus "Time to Install" seconds later. However, the good guys may rush software patches to remove the virus/worm. The "Time to Repair" is the timeafter which the computer is considered repaired -- no longer spreading the virus/worm, nor susceptible to infection again. Interpreting the results using Excel: The simulator outputs periodic counts of vulnerable, infected, and repaired computers. You can save this output to a file of type CSV (“Comma Separated Values”), which is a standard format for data files. To graph the data from a CSV file in Excel: 1. Double-click the CSV file to open it in Excel. 2. Use the mouse to select all the columns and rows that have data (or, to do this with the keyboard, press Ctrl + A). 3. From the Insert menu, click Chart. 4. Select the “XY (Scatter)” chart type, and click Finish. 5. Select File … Save As, and save the file as a *.xls file (not as a CSV file). 0200040006000800010000120000 200 400 600 800 1000 1200 Vulnerable Infected Patched Figure 2: Example chart Part 1: Worms vs. Email Viruses You may wish to read the lecture notes to refresh your memory about the difference between viruses and worms. (Your TA is happy to answer questions.) 1. Run the simulator with the settings depicted in Figure 1 above (which are the default settings), except set “Time to Install” to 10 seconds. Plot the result in an Excel chart.2. Now select “Social Network” and set “Minimum Friends” to 20. Also, change “Time to Install” to 150 seconds. Run the simulator, and plot the result in an Excel chart. 3. Which of the previous simulations is a better model for the spread of worms, and which is a better model for the spread of email viruses? Explain your answer in terms of what you know about how worms/email viruses are propagated. 4. Report the total number of computers that were infected during each simulation. To do this, look at the last line of output for the simulation, and note the number of infected and repaired computers. The total number of computers infected during the simulation is the sum of these two numbers. 5. The two totals you report in (4) should be similar to each other (if not, see the TA). Still, worms are considered more damaging than email viruses. Explain why, based on the charts of the simulations. Part 2: Vulnerability of the Network 1. Change back to the default settings (see Figure 1 above), and plot the result in an Excel chart*. 2. Now change “Initially vulnerable” to 50%, and plot the result in an Excel chart. Explain any difference between the two charts. Part 3: Infection Rate 1. Change back to the default settings, except set “Time to Spread” to 2 seconds. Plot the result in an Excel chart*. 2. Repeat Step 1 for “Time to Spread” values of 4, 6, 8 and 10 seconds. Plot the results in separate Excel charts*.
View Full Document
Unlocking...