Abstract
Several string or GUT constructions motivate the existence of a dark U(1)D gauge boson which interacts with the Standard Model only through its kinetic mixing. We compute the dark matter abundance in such scenario and the constraints in the light of the recent data from CoGENT, CDMSII and XENON100. We show in particular that a region with relatively light WIMPS, MZD≲40 GeV and a kinetic mixing 10−4≲δ≲10−3 is not yet excluded by the last experimental data and seems to give promising signals in a near future. We also compute the value of the kinetic mixing needed to explain the DAMA/CoGENT/CRESST excesses and find that for MZD≲30 GeV, δ ∼ 10−3 is sufficient to fit with the data.