1: \begin{abstract}

2: Using techniques developed recently in the field of compressed sensing

3: we prove new upper bounds for general (nonlinear) sampling numbers of (quasi-)Banach smoothness spaces in $L^2$.

4: In particular, we show that in relevant cases such as mixed and isotropic weighted Wiener classes or Sobolev spaces with mixed smoothness, sampling numbers in $L^2$ can be upper bounded by best $n$-term trigonometric widths in $\bfun$.

5: We describe a recovery procedure from $m$ function values based on $\ell^1$-minimization (basis pursuit denoising).

6: With this method, a significant gain in the rate of convergence compared to recently developed linear recovery methods is achieved.

7: In this deterministic worst-case setting we see an additional speed-up of $m^{-1/2}$ (up to log factors) compared to linear methods in case of weighted Wiener spaces.

8: For their quasi-Banach counterparts even arbitrary polynomial speed-up is possible.

9: Surprisingly, our approach allows to recover mixed smoothness Sobolev functions belonging to $S^r_pW(\T^d)$ on the $d$-torus with a logarithmically better rate of convergence than any linear method can achieve

10: when $1 < p < 2$ and $d$ is large.

11: This effect is not present for isotropic Sobolev spaces. 

12: \end{abstract}

13: