Benedicks' theorem for the Weyl transform (Q2400650)

A classical uncertainty theorem of Michael Benedicks says that if \(f\in L^1(\mathbb R^n)\) and the Lebesgue measure of the sets \[ A=\{x\in\mathbb R^n: f(x)\not=0\}\text{ and }B=\{\xi\in\mathbb R^n \hat{f}(\xi)\not =0\} \] are both finite, then \(f=0\) a.e. [\textit{M. Benedicks}, J. Math. Anal. Appl. 106, 180--183 (1985; Zbl 0576.42016)]. An analogous theorem has been established on the Heisenberg group \(\mathbb{H}^n\approx\mathbb C^n\times\mathbb R\) in terms of the finiteness of the measure of the set \(\{x\in\mathbb{H}^n:f(x)\not=0\}\) and the finiteness of the rank of the operator \(\hat{f}\), the group Fourier transform of \(f\). The geometric approach employed in the proof required the assumption that \(f\) is compactly supported [\textit{E. K. Narayanan} and \textit{P. K. Ratnakumar}, Proc. Am. Math. Soc. 138, No. 6, 2135--2140 (2010; Zbl 1202.42023)]. The article under review relaxes this assumption, in the case of the reduced Heisenberg group which is topologically \(\mathbb C\times\mathbb{T}\approx \mathbb R^2\times\mathbb{T}\) and is presented in terms of the Weyl transform of \(f\in L^1(\mathbb R^2)\): \[ (W(f)\varphi)(t)=\int _{\mathbb R^2}f(x,y)e^{i\pi(xy+2yt)}\varphi(t+x)\, dxdy. \] Theorem 1.1 of the article says that if the set \(\{w\in\mathbb R^2:f(w)\not= 0\}\) has finite Lebesgue measure and the operator \(W(f)\) has finite rank, then \(f = 0\) almost everywhere. In fact the proof is given for a reformulation of the above theorem in terms of the trace class operators on \(\mathbb{H}\), \(W(f)\) being in the trace class for \(f\in L^1(\mathbb{H})\). The approach is more along the line of arguments used by Benedicks, which yields the expected result.