$L^p$ Coarse Baum-Connes Conjecture and $K$-theory for $L^p$ Roe Algebras

Jianguo Zhang, Dapeng ZhouarXiv 2019 · DOI · arXiv

TL;DR

This paper bridges a significant gap in operator theory by verifying the $L^p$ coarse Baum-Connes conjecture for metric spaces with finite asymptotic dimension. Crucially, the authors prove that for these spaces, the $K$-theory of $L^p$ Roe algebras is independent of $p$ for all $p \in (1, \infty)$, suggesting that the underlying topological information captured by these algebras is more robust than the specific geometry of the $L^p$ space used to define them.

What Problem Does This Paper Address?

In the study of manifolds and groups, the Baum-Connes conjecture is a central pillar. It provides a way to compute "higher indices" of differential operators (like the Dirac operator) using topological tools called $K$-homology. Traditionally, this is done using $C^*$-algebras, which are built on Hilbert spaces ($L^2$ spaces).

However, $C^*$-algebras have limitations. For instance, certain "rigid" groups (those with property (T)) are notoriously difficult to handle in the $L^2$ setting. Researchers have turned to $L^p$ spaces and Banach algebras because they offer more flexibility. The "coarse" version of this conjecture deals with large-scale geometry of spaces rather than group actions.

The challenge addressed here is twofold:

  1. Does the coarse Baum-Connes conjecture actually hold in this more general $L^p$ setting?
  2. Does the choice of $p$ (the "shape" of the space's unit ball) change the resulting $K$-theory, or is the topological information invariant?

Methods

The authors employ several sophisticated techniques from non-commutative geometry and functional analysis:

  • Quantitative $K$-theory: Instead of looking at the entire algebra at once, the authors use a "controlled" approach. They analyze operators based on their propagation—essentially how much the operator "spreads" information across the metric space. By restricting this spread, they can build inductive arguments.
  • Mayer-Vietoris Sequences: This is a classic "divide and conquer" tool from topology. The authors develop a version for $L^p$ algebras to show that if the conjecture holds for smaller pieces of a space, it holds for the whole thing.
  • Riesz-Thorin Interpolation: This is the "bridge" between different $p$ values. It allows the authors to relate the $L^p$ localization algebras to the well-understood $L^2$ ($C^*$) case.
  • Anti-Čech Systems: To handle general metric spaces, they approximate them using a sequence of simplicial complexes (shapes made of triangles, tetrahedra, etc.) that capture the space's large-scale structure.

Datasets

Not discussed in the paper. As a work of pure mathematics, it relies on theoretical proofs rather than empirical datasets.

Key Insights and Ideas

  • The Power of Finite Asymptotic Dimension: The authors focus on spaces with "finite asymptotic dimension" (a concept introduced by Gromov). These spaces behave nicely at large scales, allowing for an inductive proof over the dimension of the space's "skeleton."
  • Topological Robustness ($p$-independence): The most striking "aha moment" is that for these geometric spaces, the $K$-theory doesn't care if you are working in $L^{1.5}$, $L^2$, or $L^{10}$. This suggests that $K$-theory is capturing a deep, intrinsic property of the space that transcends the specific functional analytic framework.
  • Banach vs. $C^*$ flexibility: The paper reinforces the idea that Banach algebras ($L^p$) are a viable and sometimes superior alternative to $C^*$-algebras for dealing with geometric and group-theoretic problems that are "too rigid" for Hilbert spaces.

Key Results

  • Theorem 4.6: The $L^p$ coarse Baum-Connes conjecture is verified for all proper metric spaces with finite asymptotic dimension for $p \in [1, \infty)$.
  • Theorem 5.22: For these same spaces, the $K$-theory groups $K_*(B^p(X))$ are isomorphic for all $p \in (1, \infty)$. In other words, the $K$-theory is independent of $p$.
  • Generality: Notably, the authors emphasize that their results do not require the condition of bounded geometry, making them applicable to a wider range of mathematical spaces.

Limitations and Future Work

  • The $p=1$ Case: While the conjecture is verified for $p=1$, the $p$-independence result specifically targets $p \in (1, \infty)$. The $L^1$ case often behaves differently in functional analysis (e.g., lack of reflexivity), and its relationship to the other groups remains an open question.
  • Injectivity and Expanders: The authors note that while the conjecture holds for spaces with finite dimension, "expander graphs" are known counter-examples to the injectivity of the coarse assembly map in the $C^*$ case. They raise the question of whether similar counter-examples exist in the $L^p$ world.
  • Broader Classes of Spaces: They suggest extending these results to spaces with finite decomposition complexity, a broader geometric category than finite asymptotic dimension.
  • Amenable Groupoids: The authors propose developing an $L^p$ version of the "Dirac-dual-Dirac" method to tackle conjectures related to amenable groupoids.

Summary updated

Abstract

In this paper, we verify the $L^p$ coarse Baum-Connes conjecture for spaces with finite asymptotic dimension for $p\in[1,\infty)$. We also show that the $K$-theory of $L^p$ Roe algebras are independent of $p\in(1,\infty)$ for spaces with finite asymptotic dimension.
BibTeX
@misc{zhang2019l,
  author = {Jianguo Zhang and Dapeng Zhou},
  title = {{$L^p$ Coarse Baum-Connes Conjecture and $K$-theory for $L^p$ Roe Algebras}},
  year = {2019},
  eprint = {1909.08712},
  archivePrefix = {arXiv},
  url = {https://arxiv.org/abs/1909.08712},
}

Loading related content...

Figure