Abstract:In this note, we establish a qualitative total variation version of Breuer--Major Central Limit Theorem for a sequence of the type $\frac{1}{\sqrt{n}} \sum_{1\leq k \leq n} f(X_k)$, where $(X_k)_{k\ge 1}$ is a centered stationary Gaussian process, under the hypothesis that the function $f$ has Hermite rank $d \geq 1$ and belongs to the Malliavin space $\mathbb D^{1,2}$. This result in particular extends the recent works of [NNP21], where a quantitative version of this result was obtained under the assumption that the function $f$ has Hermite rank $d= 2$ and belongs to the Malliavin space $\mathbb D^{1,4}$. We thus weaken the $\mathbb D^{1,4}$ integrability assumption to $\mathbb D^{1,2}$ and remove the restriction on the Hermite rank of the base function. While our method is still based on Malliavin calculus, we exploit a particular instance of Malliavin gradient called the sharp operator, which reduces the desired convergence in total variation to the convergence in distribution of a bidimensional Breuer--Major type sequence.
What problem does this paper attempt to address?
The problem that this paper attempts to solve is the total variation convergence problem of the sequence \(S_n(f)=\frac{1}{\sqrt{n}}\sum_{k = 1}^n f(X_k)\) under more general conditions, that is, when the function \(f\) belongs to the Malliavin - Sobolev space \(D^{1,2}\) and has an arbitrary Hermite rank \(d\geq1\). Specifically, the goal of the paper is to prove that under these conditions, \(S_n(f)\) converges to the standard normal distribution \(N(0,1)\) in the total variation distance.
### Background and Motivation
The central limit theorem (CLT) of Breuer and Major has already given the result that the sequence \(S_n(f)\) converges to the normal distribution under certain conditions. However, these conditions usually require the function \(f\) to have higher regularity (for example, belonging to the \(D^{1,4}\) space) and a specific Hermite rank (for example, \(d = 2\)). Some recent research works have further strengthened these results, proved the convergence in the total variation distance, and given quantitative convergence rate estimates.
### Main Contributions of the Paper
The main contributions of this paper lie in relaxing these conditions, in particular:
1. **Relaxing the Regularity Condition**: Reducing the required regularity of the function \(f\) from \(D^{1,4}\) to \(D^{1,2}\).
2. **Removing the Restriction on Hermite Rank**: No longer requiring the Hermite rank of the function \(f\) to be 2, but allowing any \(d\geq1\).
### Main Results
The main results of the paper can be summarized in the following theorem:
**Theorem 3**: Suppose \(f\in D^{1,2}(R,\gamma)\) and the Hermite rank \(d\geq1\), and the correlation function \(\rho\in\ell^d(N)\), that is, \(\sum_{k\in N}|\rho(k)|^d<+\infty\), and at the same time the variance \(\sigma^2\) in Theorem 1 is positive. Then, as \(n\) tends to infinity,
\[d_{TV}\left(\frac{S_n(f)}{\sqrt{\text{var}(S_n(f))}}, N(0,1)\right)\to0.\]
### Proof Methods
To prove the above results, the author uses the Malliavin - Stein method and introduces a tool called "sharp gradient". Through this tool, the author transforms the total variation convergence problem into two relatively simple steps:
1. **Two - Dimensional Breuer - Major Theorem**: Prove a two - dimensional version of the classical Breuer - Major theorem (that is, the convergence in the sense of distribution).
2. **Uniform Integrability Estimates**: Use some basic uniform integrability estimates to promote the probability convergence to \(L^1\) convergence.
### Conclusion
Through these methods, the author successfully proves that under more general conditions, the sequence \(S_n(f)\) converges to the standard normal distribution in the total variation distance. This result not only extends the existing theory, but also provides new tools and methods for future related research.