Mean-dispersion principles and the Wigner transform

Chiara Boiti,David Jornet,Alessandro Oliaro
DOI: https://doi.org/10.1007/s12220-024-01601-0
2023-04-14
Abstract:Given a function $f\in L^2(\mathbb R)$, we consider means and variances associated to $f$ and its Fourier transform $\hat{f}$, and explore their relations with the Wigner transform $W(f)$, obtaining a simple new proof of Shapiro's mean-dispersion principle. Uncertainty principles for orthonormal sequences in $L^2(\mathbb R)$ involving linear partial differential operators with polynomial coefficients and the Wigner distribution, or different Cohen class representations, are obtained, and an extension to the case of Riesz bases is studied.
Analysis of PDEs
What problem does this paper attempt to address?
The problems that this paper attempts to solve mainly focus on the application of Uncertainty Principles in orthogonal function families, especially the Uncertainty Principles related to time - frequency analysis. Specifically, the paper explores the mean and variance related to the function \(f\in L^2(\mathbb{R})\) and its Fourier transform \(\hat{f}\), and studies the relationship between these quantities and the Wigner Transform, thus providing a simple new proof to explain Shapiro's Mean - Dispersion Principle. In addition, the paper also obtains the Uncertainty Principles for orthogonal sequences in \(L^2(\mathbb{R})\), which involve linear partial differential operators with polynomial coefficients and Wigner distributions or different Cohen Class Representations, and studies the extension of these principles in the case of Riesz Bases. ### Specific problem points include: 1. **Simplified proof of the Mean - Dispersion Principle**: - The paper provides a simplified proof method for proving Shapiro's Mean - Dispersion Principle. This principle shows that in \(L^2(\mathbb{R})\), there does not exist an infinite orthogonal sequence \(\{f_k\}_{k\in\mathbb{N}_0}\) such that all of \(\mu(f_k)\), \(\mu(\hat{f}_k)\), \(\Delta(f_k)\), and \(\Delta(\hat{f}_k)\) are bounded. 2. **Generalization of the Uncertainty Principle**: - The paper further explores the Uncertainty Principles involving mean and variance, especially in orthogonal sequences. For example, for an orthogonal sequence \(\{f_k\}_{k\in\mathbb{N}_0}\), the paper proves the following inequality: \[ \sum_{k = 0}^n(\Delta^2(f_k)+\Delta^2(\hat{f}_k)+|\mu(f_k)|^2+|\mu(\hat{f}_k)|^2)\geq(n + 1)^2 \] and the equality holds if and only if there exist complex numbers \(c_k\) such that \(|c_k|=1\) and \(f_k = c_kh_k\) (where \(h_k\) is the Hermite function). 3. **Application of the Wigner Transform**: - The paper studies the Uncertainty Principle by introducing the Wigner Transform \(W(f)\) and proposes a new Mean - Dispersion Principle. Specifically, for an orthogonal sequence \(\{f_k\}_{k\in\mathbb{N}_0}\), the paper proves that: \[ \sum_{k = 0}^n\langle\hat{L}W(f_k),W(f_k)\rangle\geq(n + 1)^2 \] where \(\hat{L}\) is a specific linear partial differential operator. This result not only simplifies the proof process but also provides more flexibility and allows for the consideration of different forms of operators. 4. **Extension of Cohen Class Representations**: - The paper also studies the application of the Uncertainty Principle in Cohen Class Representations, that is, considering other types of time - frequency representations \(Q(f)\) rather than just the classical Wigner Transform. This provides a broader basis for further research. ### Summary The main objective of the paper is to provide a concise and general method to prove and generalize the Uncertainty Principles by introducing the Wigner Transform and related operators, especially for the cases of orthogonal function families and Riesz Bases. These results are not only of great theoretical significance but also provide new tools and perspectives for fields such as signal processing and quantum mechanics in practical applications.