Kakeya sets of curves (Q817670)

The paper deals with the prolific thema of Kakeya curves and begins with an interesting survey on the subject. Recall that the famous Kakeya conjecture asks whether a subset of \({\mathbb R}^n\) that includes a unit line segment in every direction must necessarily have dimension \(n\). The paper investigates the analogue problem for curves of a specified family (instead of lines). The family of curves considered by L. Wisewell is motivated by the study of one variable coefficient oscillatory integral operator, due to \textit{L. Hörmander} [Ark. Mat. 11, 1--11 (1973; Zbl 0254.42010)]: \[ T_Nf(x)=\int_{\mathbb R^{N-1}}e^{in\varphi(x,y)}a(x,y)f(y)dy \] where \(x\in{\mathbb R}^{n-1}\), \(a\in C_c^\infty\), \(\varphi\) is real-valued, smooth, with a nondegenerate condition. The question is to know whether \(T_N\) is bounded from \(L^r\) to \(L^s\), with norm \(O(N^{-n/s})\) for \(s>2n/n-1\) and \(r'\leq\frac{n-1}{n+1}s.\) Some counterexamples were given by \textit{J. Bourgain} [Geom. Funct. Anal. 1, No. 4, 321--374 (1991; Zbl 0756.42013)], who clarified the link with the initial problem on curves: \[ \begin{multlined} \text{Oscillatory integral estimates} \\ \Longrightarrow \text{Kakeya maximal function estimates} \\ \Longrightarrow\text{Kakeya sets have large dimension}.\end{multlined} \] Since 1991, there has been much progress on the straight line Kakeya problems with contributions from Bourgain, Wolff, Katz, Łaba, Tao and Schlag. There are many (absolutely nontrivial) results in the paper. It is divided into six sections: 1. Introduction, 2. ``Trivial'' results for most curves, 3. Negative results for quadratic curves, 4. Geometric methods, 5. Arithmetics methods, 6. Final remarks. The paper is very well-written and particularly rich and interesting.