Television channel spectrum sensing using independent component analysis

Spectrum reuse has been proposed as a method to solve frequency bands scarcity. Detection of unused channels in a given frequency band is the main problem to reuse the spectrum. Various methods are proposed in the literature to perform this spectrum sensing. In this study, the authors apply independent component analysis (ICA) to estimate the available bands in the TV band spectrum when analogue and digital TV channels are in operation. Each component recovered by ICA is in a specific frequency band, then the other bands are considered free to reuse, that is, the absence of transmitters in a channel means there are unused bands. A measurement setup was implemented to validate the use of ICA as a spectrum detection method. The results also show the feasibility of using ICA for wideband spectrum sensing with low computational costs.