Performance Evaluation of Computer Systems and Networks - PROJECT
Consider a 2D floorplan with N users randomly dropped in it. A random user within the floorplan produces a message, which should ideally reach all the users as soon as possible. Communications are slotted, meaning that on each slot a user may or may not relay the message, and a message occupies an entire slot. A broadcast radius R is defined, so that every receiver who is within a radius R from the transmitter will receive the message, and no other user will hear it. A user that receives more than one message in the same slot will not be able to decode it (collision).
Users relay the message they receive once, according to the following policy (trickle relaying): after the user successfully receives a message, it waits for a time window of T slots. If during that time window it correctly receives less than m copies of the same message, it relays it, otherwise it stops.
A sender does not know (or cares about) whether or not its message has been received by its neighbors.
Measure at least the broadcast time for a message in the entire floorplan, the percentage of covered users, the number of collisions.
In all cases, it is up to the team to calibrate the scenarios so that meaningful results are obtained.
Project deliverables:
- Documentation (according to the standards set during the lectures)
- Simulator code
- Presentation (up to 10 slides maximum)