A uniqueness theorem for automorphisms of a nondegenerate surface in a complex space (Q752239): Difference between revisions

Beginning with the work of H. Poincaré on local automorphisms of \(S^ 3\) in \({\mathbb{C}}^ 2\) there has been interest in studying local automorphisms of real CR-submanifolds of \({\mathbb{C}}^ n\). In the general case one considers a real analytic CR-submanifold M of \({\mathbb{C}}^ n\) of real codimension \(k\geq 1\), i.e. M is defined locally by k real analytic real-valued functions. Of interest is the group G(M) of all holomorphic maps each defined and invertible in its own neighborhood (in \({\mathbb{C}}^ n)\) of a particular point \(\xi\in M\) and which stabilize M and fix the point \(\xi\). For hypersurfaces, i.e., \(k=1\), this problem was studied by \textit{N. Tanaka} [J. Math. Soc. Jap. 14, 397-429 (1962; Zbl 0113.063)]. In order to understand the situation when \(k>1\), one has to find a suitably simple form for the submanifold M analogous to the normal form of \textit{S. S. Chern} and \textit{J. K. Moser} [Acta Math. 133, 219-271 (1975; Zbl 0302.32015)] in the case of real analytic hypersurfaces. For \(k=2\) there is the work of \textit{A. V. Loboda} [Math. USSR, Izv. 33, No.2, 295-315 (1989); translation from Izv. Akad. Nauk SSSR, Ser. Mat. 52, No.5, 970- 990 (1988; Zbl 0662.32018)], where generic real analytic submanifolds of codimension 2 in \({\mathbb{C}}^ 4\), i.e. \(n=4\) and \(k=2\), are considered. The author also looked at the case of general n and k in another work [Math. USSR, Izv. 32, No.2, 443-448 (1989); translation from Izv. Akad. Nauk SSSR, Ser. Math. 52, No.2, 437-442 (1988; Zbl 0647.32010)] and proved that the dimension of G(M) is finite by comparing its dimension with that of a certain group of a quadric. One should also consult the work of \textit{A. E. Tumanov}, in particular Math. USSR, 32, No.3, 655-662 (1989); translation from Izv. Akad. Nauk SSSR, Ser. Mat. 52, No.3, 651- 659 (1988; Zbl 0655.32026), where it is shown that infinitesimal CR- automorphisms are polynomial, but no bound is given for the degree. The present work also deals with the case \(k>1\). The author formulates the definition of one particular notion of nondegeneracy of the Levi form of M and gives some examples to illustrate this notion. Under the assumption that the Levi form of M is nondegenerate at the point \(\xi\), it is shown that every \(h\in G(M)\) is uniquely determined by its first and second derivatives at the point \(\xi\). In particular, this yields bounds on the dimension of the group G(M). Along the way a ``normal form'' for M is developed, as was alluded to above, but this is not quite so simple as in the case of hypersurfaces.