Asymmetric uniqueness sets in $\ell^q$

This paper demonstrates an asymmetry in $\ell^q$ uniqueness sets by constructing examples that fail to support measures with $\ell^q$-summable Fourier coefficients while simultaneously supporting measures with rapidly decaying positive frequencies, thereby revealing a fundamental divergence between unilateral and bilateral uniqueness problems.

The Gist

Imagine you are a detective trying to solve a mystery about hidden patterns. In the world of mathematics, specifically in a field called Harmonic Analysis, these patterns are often found in "waves" or "signals."

This paper, written by Adem Limani and Tomas Persson, is about a strange and surprising discovery: Some shapes are "one-sided" mysteries. They look completely different depending on which direction you look at them.

Here is the story of their discovery, explained without the heavy math jargon.

The Setting: The Circle and the Waves

Imagine a clock face (a circle). On this clock, we can draw different "waves" that go around it.

• Bilateral Waves: These are waves that spin both clockwise and counter-clockwise.
• Unilateral Waves: These are waves that only spin clockwise.

Mathematicians have a rulebook called $\ell^q$. Think of this rulebook as a "smoothness test." If a wave passes the test, its ripples die down very quickly as you go further out. If the ripples die down slowly, the wave is "rough" or "noisy."

The Big Question

For a long time, mathematicians believed that if a shape (a set of points on the clock) could hide a "smooth" wave spinning in one direction, it must also be able to hide a "smooth" wave spinning in both directions. They thought the rules were symmetrical.

Limani and Persson proved this wrong. They found shapes that are asymmetric.

The Two Main Discoveries

1. The "Silent Clock" (Theorem 1.1)

Imagine a clock face where almost the entire surface is painted black, except for a tiny, tiny sliver of white.

• The Trick: The authors built a specific shape (a set of points) that is so large it covers almost the whole clock (99.9% of it).
• The Asymmetry:
  • One Way (Unilateral): You can find a wave on this shape that is incredibly smooth when spinning clockwise. Its ripples vanish so fast they disappear almost instantly. It's like a whisper that stops before it reaches your ear.
  • The Other Way (Bilateral): However, if you try to find a wave that spins both ways (clockwise and counter-clockwise) on this same shape, it is impossible to make it smooth. Any wave you try to create will be incredibly noisy and rough.
• The Metaphor: It's like a room that is perfectly soundproof for sound coming from the left, but if you try to make sound travel in a circle around the room, it becomes a deafening roar. The room treats "one-way" sound and "two-way" sound completely differently.

2. The "Rough but Fast" Shape (Theorem 1.2)

They also found the opposite kind of shape.

• The Trick: They built a shape that is huge (again, almost the whole clock).
• The Asymmetry:
  • Two-Way: You can find a wave that spins both ways and is very smooth (it dies down very fast).
  • One-Way: But, if you try to make a wave that spins only clockwise and follows a specific, strict rule of smoothness, the shape refuses to let it exist. The only wave that fits is a flat line (zero).
• The Metaphor: Imagine a dance floor. You can have a couple dance perfectly together (two-way), but if you try to have a solo dancer follow a very specific, strict rhythm (one-way), the floor suddenly becomes slippery and the dancer falls. The floor supports the couple but rejects the soloist.

Why Does This Matter?

Before this paper, mathematicians thought the "rules of the game" were the same whether you looked at the waves from one side or both sides. They thought if a shape was "good" for one type of wave, it was "good" for the other.

This paper shows that nature is not always fair or symmetrical.

• You can have a shape that is "friendly" to one-sided waves but "hostile" to two-sided waves.
• You can have a shape that is "friendly" to two-sided waves but "hostile" to one-sided waves.

The "Magic Tool" They Used

To build these weird shapes, the authors used a construction method that is like building a wall with very specific gaps.

1. They started with a full circle.
2. They cut out tiny, specific gaps (arcs) in a very precise pattern.
3. They made sure the gaps were spaced out in a way that creates "traffic jams" for certain types of waves (forcing them to be noisy) but allows other types of waves to flow through smoothly.

They also used a concept called Entropy (a measure of disorder). They managed to build these shapes so that they are "ordered" enough to allow the smooth one-sided waves, but "disordered" enough to block the smooth two-sided waves.

The Takeaway

In the world of mathematics, this is a huge deal. It means that unilateral (one-sided) and bilateral (two-sided) problems are fundamentally different beasts. You cannot solve one by just assuming the rules of the other apply.

In simple terms:
Imagine you have a lock. For a long time, people thought if a key worked from the left side, it would work from the right side too. Limani and Persson built a lock where a key works perfectly from the left, but if you try to turn it from the right, the lock jams. It's a new kind of lock that breaks our intuition about how symmetry works.

Technical Summary

Here is a detailed technical summary of the paper "Asymmetric uniqueness sets in $\ell_q$" by Adem Limani and Tomas Persson.

1. Problem Statement and Motivation

The paper investigates the asymmetry between unilateral (one-sided) and bilateral (two-sided) Fourier decay for measures supported on compact subsets of the unit circle $\mathbb{T}$.

• Context: In harmonic analysis, a set $E \subset \mathbb{T}$ is a uniqueness set for a class of functions (or measures) if the only measure supported on $E$ with Fourier coefficients in that class is the zero measure.
• The Gap: Classical results (e.g., Rajchman's theorem) suggest that if Fourier coefficients decay to zero on one side, they decay on both sides. However, quantitative refinements regarding specific decay rates (like $\ell_q$ summability) are less understood.
• The Core Question: Can a set $E$ support a measure with rapid unilateral decay (e.g., super-polynomial decay for $n>0$) while simultaneously forbidding any non-trivial measure with bilateral $\ell_q$ summability (for $1 < q < 2$)?
• Significance: Previous constructions of uniqueness sets (Newman, Katznelson) often lacked quantitative control over the geometric entropy of the sets. This paper aims to construct sets with arbitrarily large Lebesgue measure (close to 1) that exhibit this extreme asymmetry, challenging the intuition that large sets must support "smooth" bilateral measures.

2. Key Definitions and Framework

• Spaces:
  • $M(\mathbb{T})$: Complex finite Borel measures.
  • $A_p(\mathbb{T})$: The space of distributions with Fourier coefficients in $\ell_p$.
  • $\ell_q$-uniqueness: A set $E$ is a uniqueness set for $\ell_q$ if no non-trivial $\mu \in M(E)$ satisfies $\sum |\hat{\mu}(n)|^q < \infty$.
• Beurling–Carleson Entropy: A geometric condition on the complement of a set $E$. If the sum of lengths of gaps times their logarithmic reciprocals is finite ($\sum |I_j| \log(1/|I_j|) < \infty$), the set supports measures with smooth Cauchy transforms (super-polynomial unilateral decay).
• Fourier Capacity: A new metric characterization introduced in the paper:
  $$\text{Cap}_{\ell^p}(E) = \inf \{ \|\phi\|_{A_p} : \phi \in C(\mathbb{T}), \phi|_E = 1 \}$$
  where $p = q/(q-1)$. The paper proves $E$ is an $\ell_q$-uniqueness set iff $\text{Cap}_{\ell^p}(E) = 0$.

3. Methodology

The authors employ a constructive approach combining three main ingredients:

1. Explicit Frequency Blocks:

   • They construct "building blocks" using convolutions of indicator functions of small arcs, denoted $\phi_{N, \delta, l}$.
   • These functions have Fourier coefficients that decay rapidly but are supported on specific arithmetic progressions (multiples of $N$).
   • By choosing parameters carefully, they force any measure supported on the constructed set to accumulate mass in specific frequency blocks, leading to $\ell_q$ divergence.

2. Arithmetic Separation (Combinatorial Lemma):

   • To ensure the frequency blocks associated with different construction stages do not overlap, the authors use a combinatorial lemma (Lemma 3.2).
   • They select integers $\{N_j\}$ inductively such that the sets of frequencies $\Lambda_j = \{ n N_j \}$ are pairwise disjoint.
   • Crucially, they control the growth of $N_j$ to ensure the resulting set $E$ retains finite Beurling–Carleson entropy.

3. Duality and Approximation:

   • The proof of the uniqueness property relies on a duality argument (Theorem 1.3). If a set has zero $\ell^p$-capacity, continuous functions vanishing on the set are dense in $A_p$.
   • This allows the construction of approximating sequences (trigonometric polynomials) that converge to 1 in the $A_p$ norm while vanishing on $E$, proving that no non-trivial measure in $\ell_q$ can exist on $E$.

4. Main Results

Theorem 1.1: The Asymmetry Phenomenon

There exists a compact set $E \subset \mathbb{T}$ with Lebesgue measure arbitrarily close to 1 such that:

1. Bilateral Divergence: Every non-trivial measure $\mu$ supported on $E$ has Fourier coefficients that are not in $\ell_q$ for any $0 < q < 2$(i.e.,$\sum |\hat{\mu}(n)|^q = \infty$).
2. Unilateral Smoothness: There exists a function $f \in L^\infty(\mathbb{T})$ supported on $E$ whose positive Fourier coefficients decay faster than any polynomial ($|\hat{f}(n)| \leq C_M n^{-M}$ for all $M$).

Significance: This shows that one can approach the $\ell_2$ threshold (where Parseval's identity forces $\ell_2$ membership for positive measure sets) arbitrarily closely while maintaining a stark divergence between unilateral and bilateral behavior.

Theorem 1.2: The Complementary Phenomenon

For any prescribed decay rate $\Omega(n) \to 0$, there exists a set $E$ of measure close to 1 such that:

1. Bilateral Near-$\ell_1$: $E$ supports a non-negative function $f \in L^2$ with Fourier coefficients in $\bigcap_{r>1} \ell_r$ (arbitrarily close to $\ell_1$).
2. Unilateral Rigidity: No non-trivial measure on $E$ satisfies the uniform unilateral bound $|\hat{\mu}(n)| \leq \Omega(n)$.

Theorem 1.3: Structural Characterization

A compact set $E$ is an $\ell_q$-uniqueness set if and only if its $\ell^p$-Fourier capacity is zero ($p = q/(q-1)$). This provides a metric characterization of uniqueness sets, linking them to the density of continuous functions vanishing on $E$ in the $A_p$ space.

Corollary 1.4: Approximation Discrepancy

The constructed sets exhibit a discrepancy between simultaneous approximation by trigonometric polynomials (which is possible) and analytic polynomials (which is rigid). Specifically, one can approximate any continuous function on $E$ by trigonometric polynomials in the $A_p$ norm, but analytic polynomials vanishing on $E$ in the $A_p$ norm must be identically zero if they approximate a function in $H^2$.

5. Extensions and Non-Periodic Analogues

The authors extend these results to the real line $\mathbb{R}$ (Section 5).

• Theorem 5.1 & 5.2: They construct compact sets $E \subset \mathbb{R}$ with positive measure exhibiting the same asymmetry for the continuous Fourier transform.
• Method: They utilize a lemma by J.-P. Kahane to transfer the discrete decay properties from the periodic setting ($\mathbb{T}$) to the continuous setting ($\mathbb{R}$).

6. Significance and Impact

1. Refutation of Entropy Necessity: The paper disproves the notion that $\ell_q$-uniqueness sets must necessarily have infinite Beurling–Carleson entropy. The authors construct sets with finite entropy (and thus large measure) that are still uniqueness sets for $\ell_q$ ($q<2$).
2. Resolution of Asymmetry: It provides a definitive answer to the question of whether unilateral spectral decay implies bilateral decay in the context of $\ell_q$ spaces. The answer is no; the two properties are fundamentally decoupled for sets of large measure.
3. New Tools: The introduction of the $\ell^p$-Fourier capacity offers a new structural tool for analyzing uniqueness sets, bridging potential theory and harmonic analysis.
4. Optimization: The construction is explicit and quantitative, allowing for precise control over the trade-off between the measure of the set, its entropy, and the decay rates of the Fourier coefficients.

In summary, Limani and Persson demonstrate that the geometry of a set can be tuned to allow for "perfect" one-sided smoothness while completely destroying two-sided summability, revealing a deep and previously unquantified asymmetry in Fourier analysis.