The diffusion of wireless sensor networks has allowed the development of a plethora of indoor localization algorithms based on these technologies. Also if several radio signal features have been exploited in order to estimate the position, the Received Signal Strength (RSS) is probably the most used. RSS depends, in addition to the distance, also on multipath transmission, barriers and non-idealities of antenna. Differences between ideal and real omnidirectional transmission patterns or polarization angle misalignment can strongly affect the RSS value impairing the following localization algorithm. In this paper, an algorithm to compensate the dependence of the RSS on the angle among the antennas is proposed and tested. The experimental results prove the goodness of the approach and the possibility of using this algorithm to minimize the dependence of RSS from the tilting angle among the nodes of a localization sensor network.