\(C^ k\)-conjugacy of holomorphic flows near a singularity (Q579674): Difference between revisions

Let X be a germ of a holomorphic vector field on \({\mathbb{C}}^ n\) vanishing at the origin, and let S be a diagonalizable element of \(g\ell (n, {\mathbb{C}})\). X is said to be an S-vector field if the linear part of X at 0 is \(S+N\) where N is a nilpotent matrix. An S-normal form is defined to be an S-vector field of the form \(S+N\), where N is a polynomial vector field which commutes with S and such that the linear part of N is nilpotent. The matrix S is said to be in the Poincaré domain if the convex hull of its eigenvalues in \({\mathbb{C}}\) does not contain 0. The point of departure of this work is the following result of Poincaré-Dulac: If S is in the Poincaré domain, then (i) there is at most a finite-dimensional set of S-normal forms; and (ii) every S-vector field is conjugate to an S-normal form. In this context, the complement of the Poincaré domain in \(g\ell (n, {\mathbb{C}})\) will be called the Siegel domain. Siegel showed that this domain contains a subset \({\mathcal S}\) of full measure such that if \(S\in {\mathcal S}\), then every S-vector field is holomorphically conjugate to S (and is thus linearizable). Holomorphic conjugacy, however, is a tricky condition and difficult to work with, and the present paper works with \(C^ k\)-conjugacy. A matrix S is called hyperbolic if its eigenvalues are simple and any two of them are \({\mathbb{R}}\)-independent. The first main result is: If S is hyperbolic, every S-vector field is \(C^ 0\)-conjugate to S (and thus is linearizable). The second main theorem covers the case where S is weakly hyperbolic (i.e., the closed line segment between two eigenvalues of S never contains 0): If S is weakly hyperbolic, then for any positive integer k, each S-vector field is \(C^ k\)-conjugate to an S-normal form.