Modelling the Faint Radio Sky: The Pathway to SKA

In the last decade our understanding of the faint radio sky has undergone a significant change. Our perception of weak radio sources has been deeply changed due to the discovery that the so-called radio-quiet active galactic nuclei are not, in fact, radio silent, but simply faint as they finally started to be detected. Moreover, a new, unexpected class of radio-loud active galactic nuclei has been recently discovered, the so-called infrared-faint radio sources. This class of rare objects discovered in deep radio surveys, is emerging as a potentially promising class of sources to pre-select high-redshift powerful radio-loud active galactic nuclei. In this Thesis, I present my contribution to the ongoing efforts to understand the nature of both these types of active galactic nuclei. The future, planned sub-mJy radio surveys (the ones expected to be delivered by the new facilities like ASKAP and SKA) will sample active galactic nuclei in the million, and the above mentioned classes will constitute a relevant fraction of the future radio catalogues. Developing methods to recognise them, and to correctly address their nature, is fundamental to fully exploiting these surveys.