Spectra related to the length spectrum (Q2143451)

The notion of Length Spectrum (LS in short) of a Riemannian manifold corresponds to the set of lengths of closed geodesics listed in ascending order. \textit{H. Huber} showed in [Math. Ann. 138, 1--26 (1959; Zbl 0089.06101); Math. Ann. 142, 385--398 (1961; Zbl 0094.05703); Math. Ann. 143, 463--464 (1961; Zbl 0101.05702)], that for compact Riemann surfaces the length spectrum and the eigenvalue spectrum are equivalent geometric quantities. Huber's result can also be obtained by Selberg trace formula, for details, see e.g. the excellent book by \textit{P. Buser} [Geometry and spectra of compact Riemann surfaces. Reprint of the 1992 original. Boston, MA: Birkhäuser (2010; Zbl 1239.32001)]. By introducing the Minimum Length Spectrum (MLS) as the set of lengths of curves that are shortest in their free homotopy classes, first \textit{C. S. Gordon} [Contemp. Math. 51, 63--80 (1986; Zbl 0591.53042)] and later \textit{C. Sormani} and \textit{G. Wei} [Trans. Am. Math. Soc. 353, No. 9, 3585--3602 (2001; Zbl 1005.53035)], showed that in a compact Riemannian manifold one has MLS\(\subset\)LS. Gordon considered multiplicity as the number of distinct free homotopy classes and showed that there are isospectral manifolds with distinct MLS. Further, Sormoni-Wei introduced the concept of the Covering Spectrum (CS), that is a subset of 1/2 MSL. The existence of isospectral compact Riemannian manifolds with different CS, for dimensions \(\geq 3\), was established \textit{B. de Smit} et al. [J. Differ. Geom. 86, No. 3, 501--538 (2010; Zbl 1219.53040)]. In the paper under review the author introduces the concepts of Extended Covering Spectrum (ECS) and Entourage Spectrum (ES) that are new metric invariants related to LS for Riemannian manifolds. These notions extend previous Sormani-Wei concepts, among them the covering spectrum CS. The author then shows that, for Riemannian manifolds M of dimension at least 3, his concept of entourage covers can topologically be characterized as those covers corresponding to subgroups of \(\pi_1(M)\) that are the normal closures of finite subsets, see Theorems 5 and 8. Further, he shows that CS\(\subset\)ES\(\subset\)MLS. These inclusions can be strict for general Riemannian manifolds. Self-similar fractals such as the Sierpinski Gasket and Carpet are also included in his approach. Two new characterizations of CS for compact geodesic spaces are given in the paper. The later notion, introduced by Sormoni-Wei as already mentioned above, is handled in the paper under review by considering a different approach, namely, through a variation of discrete homotopy methods presented by \textit{V. Berestovskii} and \textit{C. Plaut} [Topology Appl. 114, No. 2, 141--186 (2001; Zbl 0982.22002)] and \textit{C. Plaut} and \textit{J. Wilkins} [J. Topol. Anal. 8, No. 1, 89--115 (2016; Zbl 1346.57007)] respectively. Thus, instead of making use of a general construction, contained in Spanier's classical book on Algebraic Topology, as Sormoni-Wei have done in their investigation on locally path connected length spaces, the author makes use of his own previous works, with Berestovskii and Wilkins, leading to an equivalent characterization of CS, what allows the author to get further extensions, see Remark 61 in the paper under review. The main point consists of handling the new class of covering spaces, entourage covers, which is larger than the ones considered in previous papers. This approach allows the author to obtain, through the concept of E-homotopy of E-chains, where E stands for an entourage, refinements of some results of Sormoni-Wei extending their scope. The well done detailed preparation, through basic constructions, in particular the so called Ball Continuity, Chain Normalizing and Chain Lifting Lemmas, the second being an Arzelá-Ascoli type theorem, and the corresponding adaptation of the involved topological concepts to E-chains in uniform spaces, leads to the main result of the paper, Theorem 15, which describes, for compact geodesic spaces, the inclusion relations of the different classes of spectra previously introduced. Finally, the article concludes with several concrete examples pointing out subtle differences and limitations of the author's approach with respect to the one presented by Sormoni-Wei.