Assignment

You are required to simulate simple worm propagation (like the Code Red worm) in a medium-scale network by using discrete-time simulation technique introduced in class. You can do this programming project on your own computer, with any programming language among C or java or Perl or Python, but make sure your code can run under department Eustis machine.

Assume that in an isolated network with =100,000 IP address space. The IP addresses can be treated as having value from 1 to 100,000. There are N = 1000 computers vulnerable to the worm under consideration in this network. These vulnerable computers have specific IP addresses as:

1, 2, 3,..., 10,

1001, 1002, ...., 1010,

2001, 2002, ..., 2010,

.....

In other words, each cluster of 10 computers with the consecutive IPs are vulnerable to the worm, and in every 1000 consecutive IP addresses there will be one cluster of 10 vulnerable computers (so there are 100 clusters of vulnerable computers overall).

Now the worm starts its infection within this network from 1 initially infected machine that has the IP address of 1001. At each discrete time tick, a worm-infected computer can scan IP addresses within this network (the network has IP addresses). If it finds a vulnerable computer, it infects the vulnerable computer immediately and this newly infected computer can start infecting others from the next discrete time tick (no networking delay is considered, or we can say the delay is assumed to be one time unit). Let us use I(t) to represent the number of infected computers at time tick t (t=1, 2, 3, ....). Thus initially I(0) = 1.

1). Simulate a random-scanning worm propagation (like Code Red worm)

Simulate a worm propagation with the scan rate of =3. You need to simulate the worm propagation for 3 runs in order to get the vector of the number of infected, I(t), 3 times. Each of your simulation run should end when all vulnerable machines have been infected.

a). Draw a single figure shows the I(t) from these 3 simulation runs. This figure will exhibit the statistical variance in worm propagation process, but all statistically follow the worm propagation model we introduced in class. (Hint: you can save the simulated I(t) results in a file, then use Excel to import the data and draw the figure using Excel)

b). List the time for the worm finishes infecting all vulnerable computers in each run; so you need to provide 3 time values.

2). Simulate a local-preference scanning worm propagation

The above random scanning has been used by most Internet worms, but some other worms have used local-preference scanning (such as Code Red II). You need to simulate such a scanning strategy: for an infected computer with IP value x, each time it picks an IP address y trying to infect, the value y is computed as:

(1). With probability p = 0.8, it picks value y such that y Î[x-10, x+10]

(2). With the remaining probability 0.2, it picks a random IP value y between 1 to 100,000.

Again simulate the worm propagation with =3 for 3 runs, then

a). Draw one figure showing the I(t) of these 3 simulation runs.

b). List the time for the worm finishes infecting all vulnerable computers in each run; so you need to provide 3 time values.