$H^\infty$--functional calculus for generators of semigroups that admit lower bounds

This paper establishes the boundedness of $H^\infty$ functional calculi for the generators of $C_0$-semigroups on UMD Banach spaces that admit a lower bound by using a dilation argument to embed the semigroup into a $C_0$-group.

The Gist

Imagine you are a conductor of a massive, complex orchestra. Each musician represents a mathematical "operator," and the music they play represents a "semigroup"—a system that evolves or changes over time (like heat spreading through a metal rod or a population growing).

In the world of high-level math, we want to know if we can "control" this orchestra. Can we apply a specific "filter" or "musical score" (called a functional calculus) to the whole system to predict exactly how it will behave under different conditions?

This paper, written by Haak and Kunstmann, solves a specific problem: How do we control an orchestra that is moving in only one direction?

Here is the breakdown of their discovery using everyday analogies.

1. The Problem: The One-Way Street (Semigroups)

Most mathematical tools are designed for "Groups." A Group is like a reversible video: you can play it forward, or you can hit rewind and go back to exactly where you started. These are easy to study because they are balanced.

A Semigroup, however, is like a one-way street or a movie that only plays forward. Once time moves, you can't go back. Because you can't "rewind," it is much harder to prove that you can apply those "musical filters" (the $H^\infty$-functional calculus) to the system.

2. The Secret Ingredient: The "Floor" (Lower Bounds)

The authors focus on a special kind of one-way street. Imagine you are walking down a dark hallway. Usually, you might lose your sense of direction or even disappear into the shadows.

But what if you have a rule that says: "No matter how far you walk, you will always be at least 5 feet away from the walls"? This is a lower bound. It means the system never "collapses" or shrinks into nothingness. It stays "substantial."

The authors prove that if a system has even just one single moment where it is guaranteed to stay "substantial" (the lower bound), then the whole system becomes much more predictable and controllable.

3. The Strategy: The "Mirror Dimension" (Dilation)

How do they prove this? They use a brilliant trick called Dilation.

Imagine you have a dancer performing on a narrow tightrope (the one-way semigroup). It’s hard to study their full range of motion because they are restricted.

The authors use a mathematical technique (based on a method by a mathematician named Madani) to "expand" the tightrope into a massive, wide stage (a larger space). On this wide stage, the dancer isn't restricted anymore—they can move forward and backward. They have turned the one-way tightrope walk into a reversible dance (a Group).

Because the "stage" is much larger and more stable (specifically a UMD space, which is like a very well-constructed, sturdy floor), we can use all our existing, powerful tools to study the dancer on the big stage.

4. The Result: Bringing the Control Back

Once they have mastered the dancer on the big, reversible stage, they perform a "transference." They translate those rules and controls back down to the original tightrope walker.

The Conclusion: They proved that if your one-way system doesn't collapse (the lower bound), you can indeed apply those powerful mathematical "filters" to it. You can predict its behavior, even though you can't hit the rewind button.

Summary in a Nutshell

• The Goal: To find a way to mathematically "control" systems that only move forward in time.
• The Discovery: If the system is guaranteed to stay "solid" and not shrink to zero, we can pretend it's a reversible system by imagining it inside a larger, more stable world.
• The Payoff: This gives scientists and mathematicians a new set of tools to predict how complex, evolving systems will react to different forces.

Technical Summary

This paper, titled "H∞–functional calculus for generators of semigroups that admit lower bounds" by Bernhard H. Haak and Peer Chr. Kunstmann, addresses a fundamental problem in operator theory: determining when the negative generator of a $C_0$-semigroup admits a bounded $H^\infty$-functional calculus.

Below is a detailed technical summary of the work.

1. The Problem

In modern analysis, the existence of a bounded $H^\infty$-functional calculus for a generator $-A$ is a powerful property that allows for the definition of functions of operators, which is essential for solving evolution equations and studying harmonic analysis.

While it is well-established that generators of bounded groups on UMD (Unconditional Martingale Difference) Banach spaces admit such a calculus (via transference principles), the situation for semigroups is more complex. A classical way to study semigroups is to "dilate" them into groups. However, semigroups do not always admit group dilations unless they satisfy specific conditions. The authors seek to bridge this gap by investigating whether a simple lower bound on a single semigroup operator is sufficient to guarantee a bounded $H^\infty$-calculus for the generator.

2. Methodology

The authors employ a sophisticated dilation-based approach that combines two major theoretical frameworks:

• Madani’s Dilation Construction: They utilize a recent technique by Madani, which proves that if a single operator $T$ on a reflexive Banach space satisfies a lower bound ($\|Tx\| \geq c\|x\|$), one can construct an invertible dilation $S$ on a larger space $Y$. Crucially, this construction preserves geometric properties: if $X$ is reflexive, $Y$ is reflexive; if $X$ is UMD, $Y$ is UMD.
• Transference and Haase’s Results: Once the semigroup is embedded into a $C_0$-group on a larger UMD space $Y$, the authors apply existing results by Haase. Haase established that generators of $C_0$-groups on UMD spaces admit bounded $H^\infty$-calculi on specific regions (unions of sectors and strips).
• Transferring Estimates: The core technical maneuver is "transferring" the functional calculus estimates from the dilation group in the larger space $Y$ back to the original generator $A$ in the original space $X$ using commutator estimates and algebra homomorphisms.

3. Key Contributions and Results

A. Main Theorem (Bounded $H^\infty$-calculus)

The central result (Theorem 1.1) states:
If $(T(t))_{t \geq 0}$ is a $C_0$-semigroup on a UMD space $X$ with growth bound $\omega$, and there exists a single time $t_0 > 0$ and a constant $c > 0$ such that $\|T(t_0)x\| \geq c\|x\|$ for all $x \in X$, then the negative generator $-A$ (shifted by $\theta > \omega$) admits a bounded $H^\infty(\Sigma_\sigma)$ calculus for any sector $\sigma > \pi/2$.

B. Quantitative Lower Bounds

The paper provides a byproduct regarding the growth of the semigroup. They prove (Proposition 1.2) that a single lower bound at time $t_0$ implies that the semigroup satisfies a quantitative exponential lower bound for all $t > 0$:
$$\|T(t)x\| \geq m e^{\nu t}\|x\|$$
where $\nu$ is explicitly determined by $t_0$ and $c$.

C. Counterexamples and Geometric Obstructions

The authors provide critical insights into the limits of these theories:

• Example 3.1: Shows that strong continuity on $(0, \infty)$ does not prevent the semigroup norm from growing arbitrarily fast as $t \to 0^+$.
• Example 3.2: Demonstrates that the equivalence between lower bounds and the existence of a left-inverse semigroup (which holds in Hilbert spaces per the Batty-Geyer theorem) fails in general Banach spaces. They show this failure is rooted in the geometry of Banach spaces (specifically, the existence of non-complemented subspaces).

D. Refined Functional Calculus Regions

The authors extend the result to more complex domains. They prove (Corollary 4.2) that the generator $A$ admits a bounded $H^\infty$-calculus on regions $K_{\sigma, a, -\theta}$ which are unions of sectors and half-planes, providing a more precise description of the operator's functional calculus domain.

4. Significance

This paper is significant for several reasons:

1. Unification: It connects the theory of lower bounds for operators with the theory of functional calculus for semigroups.
2. Generality: It moves beyond the restrictive setting of Hilbert spaces and bounded groups, providing results for the much broader class of UMD Banach spaces.
3. Geometric Insight: By providing counterexamples, the paper clarifies that the "nice" behavior of semigroups in Hilbert spaces is often a consequence of the space's geometry rather than the semigroup properties alone.
4. Practical Utility: The quantitative exponential lower bounds and the explicit construction of the functional calculus regions provide concrete tools for researchers working on evolution equations in Banach spaces.