A cognitive radio architecture based on sub-Nyquist sampling

In this paper we discuss a cognitive radio architecture based on sub-Nyquist sampling scheme. At present, analog-to-digital converters can process wide-band signals up to 1GHz and therefore convert it into the digital baseband. To deal with the limited processing power a multicoset sampling filter in the digital domain is introduced. Thus, it is possible to have a fully flexible multiband filter for a wide-band signal with up to 1GHz bandwidth. In particular, the multiband receiver can deal with a flexible set of sub-bands with fair digital signal processing power and fast adaption.