Abstract
The aim of this paper is to study the uncertainty principles for the Dunkl wavelet transform, that set a limit to the possible concentration of a function in the time-frequency plane. Then, using for this transformation a class of concentration operators that are compact and self-adjoint, we show that their eigenfunctions are maximally time-frequency concentrated and we use its to obtain approximation inequalities for functions that are essentially concentrated in some region of the time-frequency plane.