Lower bounds for quasianalytic functions, I. How to control smooth functions?

F. Nazarov,M. Sodin,A. Volberg
DOI: https://doi.org/10.48550/arXiv.math/0208233
2003-01-20
Abstract:Consider a class of functions of one real variable with the following uniqueness property: if a function f(x) from the class vanishes on a set of positive measure, then f is the zero function. In many instances, we would like to have a quantitative version of this property, e.g. a lower bound for f(x) outside a small exceptional set. Such estimates are well-known and useful for polynomials and analytic functions. In this work we prove similar results for the Denjoy-Carleman and the Bernstein classes of quasianalytic functions.
Classical Analysis and ODEs
What problem does this paper attempt to address?
The problem that this paper attempts to solve is the quantitative uniqueness estimation problem of smooth functions in the quasi - analytic function class. Specifically, the authors hope to provide an estimate that controls the norm of a function \( f \) from below when it is restricted to a positive - measure set \( E \) for functions \( f \) in the quasi - analytic function class \( F \). Such estimates have very good results for polynomials, complex - analytic and real - analytic functions, exponential polynomials, etc., but the results for quasi - analytic function classes (such as the Denjoy - Carleman class and the Bernstein class) are still insufficient. ### Main problems of the paper 1. **Quantitative uniqueness estimation**: - For a given quasi - analytic function class \( F \), if a function \( f\in F \) takes zero values on a positive - measure set \( E \), then \( f \) must be the zero function. The authors hope to find a quantitative version of this property, that is, to provide a lower - bound estimate for \( |f| \) in the region outside \( E \). 2. **Introduction of Bang degree**: - To achieve the above purpose, the authors introduce a new index - Bang degree \( n_f \), which is used to describe the behavior of quasi - analytic smooth functions. The Bang degree depends on the growth of the sequence \( \{M_j\} \) and the lower bound of \( ||f||_{[0,1]} \). 3. **Main theorems**: - **Theorem A**: The number of zeros (including multiplicities) of any function \( f \) belonging to the \( C_A([0,1]) \) class does not exceed its Bang degree \( n_f \). - **Theorem B**: For any interval \( I\subset [0,1] \) and measurable subset \( E\subset I \), we have \[ ||f||_I \leq \left(\Gamma(2n_f)\frac{|I|}{|E|}\right)^{2n_f}||f||_E. \] where \( \Gamma(n) = 4e^4+\gamma(n) \), and \( \gamma(n) \) is a quantity related to the function \( A \). ### Methods and techniques The authors use elementary calculus techniques to prove these results and utilize Bang's method, especially when dealing with quasi - analytic function classes. By introducing the Bang degree and combining the classical Remez inequality and the Bernstein degree concept, the authors successfully provide similar estimates for quasi - analytic functions. In conclusion, the main goal of this paper is to provide quantitative uniqueness estimates for functions in the quasi - analytic function class and introduce the Bang degree as an important feature to control the behavior of these functions.