Table of Contents
Fetching ...

Characterizations of positive operators via their powers

Hranislav Stanković

TL;DR

This work develops new criteria for when an operator on a complex Hilbert space is positive or normal by inspecting its powers. It generalizes Putnam’s results through conditions on the square $T^2$, the numerical range of odd powers, and ascent/descent, showing that certain power-structure constraints force normality, and that positivity can be recovered from accretivity of powers, even under relaxed hypotheses. A key contribution is the equivalence between normality (and positivity) and the existence of coprime exponent pairs with powers possessing the respective property. The results yield practical criteria for identifying positive operators and connect accretivity of powers to the spectral and numerical-range geometry, with implications for operator theory and quantum-mechanical contexts where positivity and normality are central.

Abstract

In this paper, we present new characterizations of normal and positive operators in terms of their powers. Among other things, we show that if $T^2$ is normal, $\mathcal{W}(T^{2k+1})$ lies on one side of a line passing through the origin (possibly including some points on the line) for some $k\in\mathbb{N}$, and $\mathrm{asc\,}(T)= 1$ (or $\mathrm{dsc\,}(T)=1$), then $T$ must be normal. This complements the previous result due to Putnam [28]. Furthermore, we prove that $T$ is normal (positive) if and only if $\mathrm{asc\,}(T)= 1$ and there exist coprime numbers $p,q\geq 2$ such that $T^p$ and $T^q$ are normal (positive). Finally, we also show that $T$ is positive if and only if $T^k$ is accretive for all $k\in\mathbb{N}$, which answers the question from [22] in the affirmative.

Characterizations of positive operators via their powers

TL;DR

This work develops new criteria for when an operator on a complex Hilbert space is positive or normal by inspecting its powers. It generalizes Putnam’s results through conditions on the square , the numerical range of odd powers, and ascent/descent, showing that certain power-structure constraints force normality, and that positivity can be recovered from accretivity of powers, even under relaxed hypotheses. A key contribution is the equivalence between normality (and positivity) and the existence of coprime exponent pairs with powers possessing the respective property. The results yield practical criteria for identifying positive operators and connect accretivity of powers to the spectral and numerical-range geometry, with implications for operator theory and quantum-mechanical contexts where positivity and normality are central.

Abstract

In this paper, we present new characterizations of normal and positive operators in terms of their powers. Among other things, we show that if is normal, lies on one side of a line passing through the origin (possibly including some points on the line) for some , and (or ), then must be normal. This complements the previous result due to Putnam [28]. Furthermore, we prove that is normal (positive) if and only if and there exist coprime numbers such that and are normal (positive). Finally, we also show that is positive if and only if is accretive for all , which answers the question from [22] in the affirmative.

Paper Structure

This paper contains 5 sections, 20 theorems, 36 equations.

Key Result

Theorem 2.1

AndersonTrapp75 Let $S$ be a closed subspace of $\mathcal{H}$ and $T\in\mathfrak{B}(\mathcal{H})$ have the matrix operator decomposition induced by $S$ and given by (eq:matrix_decomposition). Then, $T$ is positive if and only if

Theorems & Definitions (30)

  • Theorem 2.1
  • Theorem 2.2
  • Theorem 2.3: Löwner-Heinz inequality Heinz51Lowner34
  • Theorem 2.4: Kato inequality Kato52
  • Theorem 2.5: Fuglede-Putnam Theorem Fuglede50Putnam51
  • Corollary 2.1
  • Corollary 3.1
  • Corollary 3.2
  • Corollary 3.3
  • Theorem 3.1
  • ...and 20 more