Till KTH:s startsida Till KTH:s startsida

Ändringar mellan två versioner

Här visas ändringar i "Project" mellan 2015-11-25 12:43 av Viktoria Fodor och 2015-11-25 15:51 av Viktoria Fodor.

Visa nästa > ändring.

Project

Project results¶ Thank you for the project submissions. There were many nice solutions. Finally, Eva Azoidou, Dinh Tuan Hai and Roberto Congiu got the 5 extra points.¶ As an example, see Hai's project.¶ Many of you submitted figures where the texts on the axes and labels were not visible on the printout. This is a very annoying, though typical mistake. Be careful with it in the future.¶ Some of you submitted surprising results, like system performance improving with number of hops or with increasing error probability. These kind of results are usually not reasonable, and show that you made some errors in your modeling or calculations.¶ Project description

There is a mandatory small project in the course with submission deadline Monday, 2015 March 23. The project has to be performed individually and is graded pass/fail based on the submitted report. Projects with exceptional quality receive 5 extra points that is added to the points achieved at the exam. These extra points are valid only at the first exam, and are not considered at make-up exams.

We offer two projects this year. You have to solve one of them. Project descriptions will be added by middle of February.

I. Modeling Error Control in Relay Networks

Project description (pdf) (updated with some more information. See highlighted text.)

II. Simulating a Media Stream Service

Project Description (pdf)

Aux: song popularities

Suggestions and typical problems¶

Characters and texts (labels, captions) in figures and graphs should be possible to read in a printout. Please make sure you use a good font size. This is a very annoying, though typical mistake. Try to avoid it!¶

Results should be reasonable. E.g., the performance of a network should not degrade if the capacity is increased, the loss probability should not decrease if there are more lossy links to cross, and so on. These kind of results are usually not reasonable, and show that you made some errors in your modeling or calculations.