Table of Contents
Fetching ...

Strengthening of spectral radius, numerical radius, and Berezin radius inequalities

Pintu Bhunia

TL;DR

The paper develops refined numerical radius and Berezin radius bounds for $n\times n$ operator matrices and RKHS operators, enabling sharper estimates for the spectral radius of sums, products, and commutators. The main technique combines a key inner-product inequality for pairs of bounded operators with a matrix-entry bound, yielding $w({\bf A}) \le w([a_{ij}])$ and $\textbf{ber}(\mathbf{A}) \le w([a_{ij}])$ with explicitly structured entries. These bounds then feed into applications including improved Kronecker-product bounds, spectral-radius estimates, and sharper bounds for the zeros of algebraic equations via the Frobenius companion matrix. The results refine and generalize several known inequalities in the literature and are illustrated with Hardy space examples, highlighting potential impact in operator theory and numerical analysis.

Abstract

Suppose $\mathcal{H}_1, \mathcal{H}_2, \ldots, \mathcal{H}_n$ are arbitrary complex Hilbert spaces, and ${\bf A}=[A_{ij}]$ is an $n\times n$ operator matrix with $A_{ij}\in \mathcal{B}(\mathcal{H}_j, \mathcal{H}_i).$ We show that $w({\bf A}) \leq w\left(\begin{bmatrix} a_{ij} \end{bmatrix}_{i,j=1}^n \right),$ where $w(\cdot)$ denotes the numerical radius and the entries $$ a_{ij}=\begin{cases} w(A_{ii}) & \textit{if $i=j$}, \sqrt{ \left( \|A_{ij}\|+\|A_{ji}\| \right)^2- \left(\|A_{ij}\| \|A_{ji}\|-w(A_{ji}A_{ij}) \right)}^{} & \textit{if $i<j$}, 0 & \textit{if $i>j$.} \end{cases}$$ This bound improves $w({\bf A}) \leq w\left(\begin{bmatrix} a'_{ij} \end{bmatrix}_{i,j=1}^n \right),$ where $a'_{ij}=w(A_{ii})$ if $i=j$ and $a'_{ij}=\|A_{ij}\|$ if $i\neq j$. We deduce an upper bound for the Kronecker products $A\otimes B$, where $A\in \mathcal{M}_n(\mathbb{C})$ and $B\in \mathcal{B}(\mathcal{H}_1)$, which refines Holbrook's classical bound $w(A\otimes B)\leq w(A)\|B\|$, when all entries of $A$ are non-negative. Further, we obtain the Berezin radius inequalities for $n\times n$ operator matrices where the entries are reproducing kernel Hilbert space operators. We provide an example, which illustrates these inequalities for some concrete operators on the Hardy--Hilbert space. Applying the numerical radius bounds, we show that if $A_i \in \mathcal{B}(\mathcal{H}_i, \mathcal{H}_1) $ and $B_i\in \mathcal{B}(\mathcal{H}_1, \mathcal{H}_i)$ for $i=1,2,$ then \begin{eqnarray*} r(A_1B_1+A_2B_2) \leq \frac{ 1 }{2 } \left(w(B_1A_1)+w(B_2A_2) \right) + \frac{ 1 }{2 } \sqrt{ \left(w(B_1A_1)-w(B_2A_2)\right)^2 + 3\|B_1A_2\|\|B_2A_1\| + η}, \end{eqnarray*} where $η=w(B_2A_1 B_1A_2)$, and $r(\cdot)$ denotes the spectral radius. We also achieve a bound for the roots of an algebraic equation.

Strengthening of spectral radius, numerical radius, and Berezin radius inequalities

TL;DR

The paper develops refined numerical radius and Berezin radius bounds for operator matrices and RKHS operators, enabling sharper estimates for the spectral radius of sums, products, and commutators. The main technique combines a key inner-product inequality for pairs of bounded operators with a matrix-entry bound, yielding and with explicitly structured entries. These bounds then feed into applications including improved Kronecker-product bounds, spectral-radius estimates, and sharper bounds for the zeros of algebraic equations via the Frobenius companion matrix. The results refine and generalize several known inequalities in the literature and are illustrated with Hardy space examples, highlighting potential impact in operator theory and numerical analysis.

Abstract

Suppose are arbitrary complex Hilbert spaces, and is an operator matrix with We show that where denotes the numerical radius and the entries This bound improves where if and if . We deduce an upper bound for the Kronecker products , where and , which refines Holbrook's classical bound , when all entries of are non-negative. Further, we obtain the Berezin radius inequalities for operator matrices where the entries are reproducing kernel Hilbert space operators. We provide an example, which illustrates these inequalities for some concrete operators on the Hardy--Hilbert space. Applying the numerical radius bounds, we show that if and for then \begin{eqnarray*} r(A_1B_1+A_2B_2) \leq \frac{ 1 }{2 } \left(w(B_1A_1)+w(B_2A_2) \right) + \frac{ 1 }{2 } \sqrt{ \left(w(B_1A_1)-w(B_2A_2)\right)^2 + 3\|B_1A_2\|\|B_2A_1\| + η}, \end{eqnarray*} where , and denotes the spectral radius. We also achieve a bound for the roots of an algebraic equation.

Paper Structure

This paper contains 5 sections, 22 theorems, 62 equations.

Key Result

Theorem 1.2

Let $\mathcal{H}_1, \mathcal{H}_2, \ldots, \mathcal{H}_n$ be complex Hilbert spaces, and let ${\bf A}=[A_{ij}]$ be an $n\times n$ operator matrix with $A_{ij}\in \mathcal{B}(\mathcal{H}_j, \mathcal{H}_i).$ Then

Theorems & Definitions (47)

  • Definition 1.1
  • Theorem 1.2
  • Remark 1.3
  • Theorem 1.4
  • Remark 1.5
  • Corollary 1.6
  • Remark 1.7
  • Theorem 1.8
  • Remark 1.9
  • Lemma 2.1: Buzano
  • ...and 37 more