Skyline supports IMS data for Waters and Agilent instruments. Using a drift time predictor can help Skyline ignore unwanted signals and improve data quality. By specifying the predicted drift time for each peptide of interest you can tell Skyline to ignore other scans that might contribute noise.
To add or change a drift time predictor, use the Settings|Peptide Settings menu item and select the Prediction tab, then the Drift Time Predictor button to bring up the Drift Time Predictor table editor.
The easiest way to set up a drift time predictor table is to start with a Skyline document with imported results, then use the "Use Results" button in the Drift TIme Predictor editor. This simply scans the existing imported results and determines the drift time of the scan containing the most intense peak. Once you have that, you can reimport the data and Skyline can ignore scans at the proper retention time but wrong drift time.
There is a risk, of course, that the most intense peak at a given retention time isn't actually that of the precursor you are interested in, in which case you will be making the noise situation worse instead of better. The ideal way to use this training feature is with simple training sets that elute one precursor at a time. If you do not have that capability then you should go through and verify the drift time selections manually using the chromatogram viewer's intensity heat map of mz vs drift time.