Time-frequency concentration and localization operators in the Dunkl setting (Q330594)

The Dunkl transform is a generalization of the usual Fourier transform to an integral transform invariant under a finite reflection group. It is relevant for the study of quantum many body systems of Calogero-Moser-Sutherland type. The exponential function in the Fourier transform is replaced by a kernel \(\mathcal K\), which is the solution of a joint eigenvalue problem for the associated Dunkl operator. It involves a weight function \(w_k\), but otherwise looks like the Fourier transform \[ \mathcal F_{D} f(x) = c_k \int_{\mathbb R^d} \mathcal K (-i\xi, x) f(x) w_k(x) dx, \] where you can consult the paper for the value of the constant and the weight function \(w_k\). A Heisenberg uncertainty inequality has been proved for it, which rules out simultaneous localization of a function and its Dunkl transform, so one studies a windowed version, the Dunkl-Gabor transform, with window \(g\) \[ \mathcal G^D_{g} f(x, \xi) = \int_{\mathbb R^d} f(s) \overline{ \tau_x \mathcal M_{\xi}g (s) } \, w_k(s) ds, \] where \[ \tau_x f := \mathcal F_{D}^{-1} [ \mathcal K(ix, \cdot) \mathcal F_{D}(f)], \] and \[ \mathcal M_{\xi} g := \mathcal F_{D} \left( \sqrt{ \tau_{\xi} |g|^2} \right). \] There is also a Heisenberg inequality for the Dunkl-Gabor transform, which is presented in two equivalent forms. The author gives a refined version of this Heisenberg inequality for orthonormal sequences, which can be summarized by saying that while the sum form of Heisenberg predicts that a sum of terms where you have substituted an orthonormal sequence will grow like \(N\), the author shows that it grows like \(N^{1 + \frac{s}{2\gamma +d}}\) , where all the term in the power are positive giving faster growth of the sum. For the values of the constants in the power, I refer to the paper. The author concludes with a Schatten class estimate for a time-frequency localization of a pseudo-differential operator, \(\mathcal A_{\mathcal S}^{g_1, g_2}\), with symbol \(\mathcal S\), analysis window function \(g_1\) and synthesis window function \(g_2\), \[ \mathcal A_{\mathcal S}^{g_1, g_2} = \int \mathcal S (x, \xi) \mathcal G_{g_1}(f)(x, \xi) g_2(x, \xi)w_k (x) w_k(\xi) dx d\xi. \]