The work is divided into three main parts. In a first time (chapter 1) we define a “BTZ” black hole in anti de Sitter space in any dimension. That will be done by means of group theoretical and symmetric spaces considerations. A physical “good domain” is identified as an open orbit of a subgroup of the isometry group of anti de Sitter. Then (chapter 2) we show that the open orbit is in fact isomorphic to a group (we introduce the notion of globally group type manifold) for which a quantization exists. The quantization of the black hole is performed and its Dirac operator is computed. The third part (appendix A and B) exposes some previously known results. Appendix A is given in a pedagogical purpose: it exposes generalities about deformation quantization and careful examples with SL(2,R), and split extensions of Heisenberg algebras. Appendix B is devoted to some classical results about homogeneous spaces and Iwasawa decompositions. Explicit decompositions are given for every algebra that will be used in the thesis. It serves to make the whole text more self contained and to fix notations. Basics of quantization by group action are given in appendix A.4. One more chapter is inserted (chapter 3). It contains two small results which have no true interest by themselves but which raise questions and call for further development. We discuss a product on the half-plane or, equivalently, on the Iwasawa subgroup of SL(2,R), due to A. Unterberger. We show that the quantization by group action machinery can be applied to this product in order to deform the dual of the Lie algebra of that Iwasawa subgroup. Although this result seems promising, we show by two examples that the product is not universal in the sense that even the product of compactly supported functions cannot be defined on AdS2 by the quantization induced by Unterberger's product. Then we show that the Iwasawa subgroup of SO(2,n) (i.e. the group which defines the singularity) is a symplectic split extension of the Iwasawa subgroup of SU(1,1) by the Iwasawa subgroup of SU(1,n). A quantization of the two latter groups being known, a quantization of SO(2,n) is in principle possible using an extension lemma. Properties of this product and the resulting quantization of AdSl were not investigated because we found a more economical way to quantize AdS4 .