While in the near future everything will be tagged with Radio Frequency IDentification (RFID) tags, the localization of these tags in their environment is becoming an important feature for many RFID-based ubiquitous computing applications and robotics. Location-aware services in RFID system will allow offering value-added services to the RFID user, and RFID tags can be used for more than just labeling items. This paper proposes an RSS-based positioning algorithm for objects attached with UHF RFID tags, by means of two mobile RFID antennas and landmarks to overcome the limitations of RFID technology and reduce the localization cost and environment complexity. The proposed algorithm opens up a possibility for creating novel location-based applications using RFID technology, without specialized hardware or extensive training. It uses an RFID map made from passive or active reference tags (landmarks) to locate analytically any unknown tag detected by the RFID Reader antennas and improve statistically the overall accuracy of locating objects by defining the statistical distribution of the location estimation error for each landmark. This algorithm is independent from the readers coordinates, and hence it can be more practical due to its mobility and its low cost to achieve a high deployment. To minimize the effect of the RSS and the process measurement noises on the position estimation, an adaptive Kalman filter and probabilistic map matching are applied. Results obtained after conducting extensive simulations demonstrate the validity and suitability of the proposed positioning algorithm to provide high-performance level in terms of accuracy and scalability.