The signal is the component of the observed data (e.g. of a time series) that carries useful information. The complementary (opposite) concept is noise.
In a narrower sense (e.g. in signal processing) signals are functions of time, as opposed to fields (functions of spatial coordinates) or images.