Hadron beams - approximately 200 MeV protons and 400 MeV/u fully stripped carbon ions - are better suited to treat deep-seated tumours than X ray beams - produced by electrons accelerated by linear accelerators (linacs) to 5-25 MeV - because they leave the maximum energy density at the end of their range in matter, in the so-called Bragg peak. This physical property allows dose depositions that are more conformal to the tumour target and spare much better the surrounding healthy tissues, so that hadron therapy treatment rooms could substitute X ray therapy rooms (about 20 000 worldwide) if the needed accelerators could be made of similar dimensions and costs. Cyclotrons and synchrotrons are at the heart of today proton and carbon ion therapy centres, respectively. At the end of 2015, more than 130 000 patients have been treated with proton beams and almost 20 000 with carbon ions [Particle Therapy Co-Operative Group data, \url{https://www.ptcog.ch/index.php/patient-statistics}]. High frequency hadron therapy linacs have been studied in the last 30 years. Their main advantage is represented by their active beam energy modulation, which permits quick treatments with superior beam quality and novel dose delivery techniques.

This thesis is the last of many research works on the development of linear accelerators for hadron therapy. The preliminary design of two linear accelerator facilities, for proton therapy (TULIP) and carbon ion therapy (CABOTO), was completed.

The introductory Chapter will review the rationale of hadron therapy as a treatment methodology, together with a discussion of the advantages of linacs in hadron therapy compared to state-of-art technologies. Finally, the most important past research activities based on which this thesis started will be presented.

The second Chapter describes the RF design of accelerating structures performed for the proton and carbon ion therapy facilities later on presented.

The third Chapter is dedicated to the proton linac design, called TULIP: TUrning LInac for Protontherapy.

The forth Chapter describes the carbon ion linac design, called CABOTO: CArbon BOoster for Therapy in Oncology.

The fifth Chapter presents the experimental activity performed on the backward travelling wave prototype built. Goal of the test is to study the high-gradient break-down limitation of S-Band cavities.

The thesis is completed by an Appendix where the beam dynamics codes developed to design the linacs are discussed.