Difference between revisions of "Private:progress-alkurbi"
From NMSL
| Line 7: | Line 7: | ||
=== Feb 01 === | === Feb 01 === | ||
| − | * | + | * Works: |
| − | ** | + | ** Exporting statistics into Matlab, and fixing figures' errors (Still in progress). |
| + | ** Generated two new 24h traffic with: one with 50 bots, and the other with 100 bots. | ||
| + | ** Calculating FP/FN for all combinations and generate 24 statistic reports. I'm waiting for Prof suggestion on this matter as calculating FP/FN while applying the same setting (at least 24 different settings) will require an intensive and very long time. | ||
| + | ** Export the 24 statistic reports into Matlab, and generate 24 figures. | ||
| + | ** Finally and to prove the ineffectiveness of picking (Win=1h), I'll need to run some experiments to calculate FP/FN as well, plot a figure and include it in the report. | ||
| + | ** Update the report. | ||
| + | |||
=== Jan 24 === | === Jan 24 === | ||
| Line 16: | Line 22: | ||
** Ran the proposed system against the generated traffic with different win sizes (3h, 2h), and different Sliding-Win sizes (5m, 10m, 15m, 20m, 25m, 30m), to calculate False Positives/Negatives, and generated 12 statistics report. | ** Ran the proposed system against the generated traffic with different win sizes (3h, 2h), and different Sliding-Win sizes (5m, 10m, 15m, 20m, 25m, 30m), to calculate False Positives/Negatives, and generated 12 statistics report. | ||
** Exporting statistics reports into Matlab and generating figures. | ** Exporting statistics reports into Matlab and generating figures. | ||
| − | |||
| − | |||
Revision as of 10:41, 28 January 2011
Spring 2011
- Courses: None
- Research: Developing Online SIP-Botnet Detection System
- Progress Report: Please read "Progress" section here
Feb 01
- Works:
- Exporting statistics into Matlab, and fixing figures' errors (Still in progress).
- Generated two new 24h traffic with: one with 50 bots, and the other with 100 bots.
- Calculating FP/FN for all combinations and generate 24 statistic reports. I'm waiting for Prof suggestion on this matter as calculating FP/FN while applying the same setting (at least 24 different settings) will require an intensive and very long time.
- Export the 24 statistic reports into Matlab, and generate 24 figures.
- Finally and to prove the ineffectiveness of picking (Win=1h), I'll need to run some experiments to calculate FP/FN as well, plot a figure and include it in the report.
- Update the report.
Jan 24
- Works:
- Generated 24h SIP traffic with "1000" users, "10" bots.
- Tuned Alpha & Beta values.
- Ran the proposed system against the generated traffic with different win sizes (3h, 2h), and different Sliding-Win sizes (5m, 10m, 15m, 20m, 25m, 30m), to calculate False Positives/Negatives, and generated 12 statistics report.
- Exporting statistics reports into Matlab and generating figures.
