Many developed techniques for indoor positioning aim at providing a mass market terminal with the continuity of the location service, from outdoors where GNSS are almost unbeatable, to indoors. The real joined constraints of real world, mass market and indoor positioning are so far too difficult and no acceptable systems have yet been deployed (fortunately for IPIN conferences!). Our current purpose is slightly different: the question we would like to answer is "what is the limit of the possible accuracy of a GNSS-like transmitter based system for indoor positioning?" The target is now the professional world where a few constraints can be acceptable: especially cost and deployment rules. Of course, one knows that optical systems are well established and that the centimeter level should be reached in order to present any interest. The goal of this first paper on this subject is to cope with the basics of this problem and to review possible theoretical aspects. Thus, differential techniques are detailed with associated requirements in order to make it possible to reach that level of accuracy. The specificity of the indoor environment is highlighted and is mainly considered through the way of increased noise levels on the various measurements. Multipath induced noise is also modeled and its drastic impact evaluated. Simulations are carried out and it is shown that the centimeter level is almost impossible to reach as soon as the measurement noise increased a little bit. As a conclusion, a few directions of future works are suggested in order to overcome the increased noise level that is almost inevitable indoors