On the Invertibility of Born-Jordan Quantization

TL;DR

This paper investigates the invertibility of Born-Jordan quantization, showing that operators can be represented in Born-Jordan form but not uniquely in general, and establishing conditions for uniqueness in specific function spaces.

Contribution

It demonstrates that any operator can be expressed in Born-Jordan form, analyzes the non-uniqueness issue, and identifies conditions for unique representation in certain function spaces.

Findings

Operators can be written in Born-Jordan form, but the representation is not unique for general symbols.

Uniqueness of the representation is established under exponential growth constraints in specific function spaces.

The paper employs distribution division techniques to analyze the invertibility of Born-Jordan quantization.

Abstract

As a consequence of the Schwartz kernel Theorem, any linear continuous operator $\widehat{A}:$ $\mathcal{S}(\mathbb{R}^{n})\longrightarrow\mathcal{S}^{\prime}(\mathbb{R}^{n})$ can be written in Weyl form in a unique way, namely it is the Weyl quantization of a unique symbol $a\in\mathcal{S}^{\prime}(\mathbb{R}^{2n})$. Hence, dequantization can always be performed, and in a unique way. Despite the importance of this topic in Quantum Mechanics and Time-frequency Analysis, the same issue for the Born-Jordan quantization seems simply unexplored, except for the case of polynomial symbols, which we also review in detail. In this paper we show that any operator $\widehat{A}$ as above can be written in Born-Jordan form, although the representation is never unique if one allows general temperate distributions as symbols. Then we consider the same problem when the space of temperate distributions is replaced by the space of smooth slowly increasing functions which extend to entire function in $\mathbb{C}^{2n}$, with a growth at most exponential in the imaginary directions. We prove again the validity of such a representation, and we determine a sharp threshold for the exponential growth under which the representation is unique. We employ techniques from the theory of division of distributions.