Compressive sensing is a new approach to digital signals acquisition and processing. Its main aim is the reconstruction of sparse signals or images (i.e. with few significant coefficients) from what was previously believed to be incomplete information. Compressive sensing has many potential applications, not only limited to the analysis of signals or images. In this talk we will first describe the main theoretical properties of compressive sensing, then three solution algorithms based on \ell_1 minimization will be presented, along with numerical experiments to highlight the main characteristics of these three methods in the framework of compressive sensing.