Résumé : While far from being technologically mature, fusion is an attractive low-carbon source ofenergy that could play an important role in the future world electricity and heat productionmix. Following this target, the next step after demonstrating a net energy gain in fusion willbe the construction of a demonstration fusion reactor to produce heat or electricity.To reach a burning state, where the energy of the fusion products is sufficient to sustainthe fusion reaction, one first needs to bring the fusion fuel above an energy threshold. Oneof the experimentally proven techniques used in present day tokamaks to heat the plasmafuel launches microwaves in the Ion Cyclotron Range of Frequencies (ICRF) in the low temperatureand density edge of fusion devices using an antenna. This heating method, calledIon Cyclotron Resonance Heating (ICRH), is cost effective, has a mature radio-frequencysource technology and comes with highly flexible and efficient heating scenarios. ICRH alsocovers other important fusion applications like current drive, turbulence stabilization in theplasma, wall conditioning, etc.For these reasons, the installation of an ICRH system is considered for future fusiondemonstration power plants like DEMO or ARC. However, future fusion reactor conditionsare quite challenging for ICRH. The two most stringent aspects of a reactor lies (i) in theexpected low density profile in the edge leading to a low coupling of power to the plasma(ii) the restricted space available for auxiliary components inside the reactor’s chamber andsurrounding wall structures. The other challenges faced are the installation, availability andmaintainability of these components in a reactor environment. On top of these challenges theICRH system should minimize the plasma wall interactions and the impurity release inducedduring its operation.Two very different ICRH systems are proposed for the next reactor generation. One relieson high power high density antennas inserted in a port of the fusion device. This solution isthe one implemented in ITER. The other relies on low power low density antenna sectionsdistributed over the periphery of the device. This solution comes with new challenges andbenefits compared to the port-plug solution.When preparing this thesis, a first comparison of the merits of the two systems was initiated.This objective first required a good understanding of wave coupling and wave propagationinside a plasma and was achieved through the use of the ANTITER antenna simulationcode. It also required to carefully follow the experiments and methods developed to minimize ICRH plasma wall interactions and related impurity release induced during ICRH operation.This study eventually led to the analysis of an edge loss mechanism that was largely neglecteduntil now. In parallel, the two systems proposed for DEMO were analyzed, first usingANTITER and then using a 3D commercial electromagnetic software.