So you just have the N1s peak recorded and you want to know what species are there, right? Follow these steps:
1) First, you need to know your component width. That is machine- and setting-specific, so you need to determine it best by recording the spectrum of a sample for which you are certain that there is only one nitrogen species included.
2) Baseline subtraction: there is quite the debate what baseline should be subtracted (http://www.casaxps.com/help_manual/manual_updates/peak_fitting_in_xps.pdf) but most people nowadays use a Shirley baseline when they want to perform a component analysis. Use the same baseline function type for both your component width reference and your actual sample.
3) Use the minimum number of functions as determined from your reference sample to reconstruct your sample spectrum. For N1s a simple Gaussian usually works fine. You can allow the width to vary a little, but I wouldn't allow more than 0.1 eV, otherwise you will either overwrite components or overparametrize your fit.
4) Since N1s does not involve any coupling as a p or d signal would, the atomic concentrations are directly proportional to your peak integrals.
For more details, refer to the Casa XPS manual I linked above. Even if you don't have Casa the procedures are the same, they just require a bit more manual work.