It's very simple. There are several ways to do it, but I suggest to use GenAlEx. Just prepare genotypes of the individuals in Excel file, then prepare GenAlEx input file using 'Template' option, and then use 'Export' option to create input file for Arlequin.
Arlequin datafile is a mess (a .xml thing not really easy to build). Hopefully a bench of softwares could be used to create it providing you have your data in a spreadsheet or notepad format (Create, PDPSpider, Convert,...). Best is probably to use PGDSpider as it will able you to convert your file to multiple formats (Create is also a good option).
http://www.cmpg.unibe.ch/software/PGDSpider/
http://bcrc.bio.umass.edu/pedigreesoftware/node/2
Coding in presence/absence is not the better way to code the microsats, best is perhaps to use the number of repeats (if you have this information) as it will allow you to compute some SSR-specific kind of analysis (i.e. assuming specific models of evolution as the Stepwise Mutation Model), especially in Arlequin which is able to compute these.
So you will avoid many problems in the completion of your research, if you do not understand what the program does, its potential and limitations, you will be going around and wasting precious time.
Before working with a package, read the manual carefully.
For the correct format of your data, you can use Transformer-4. This software allow conversion of your data in excel table to other formats used in software of genetic analysis (dominant and codominant markers).