In　order　to　improve　the　quality　of　the　RSS　(Received　Signal　Strength)　during　the　offline　phase,　a　Mixture　Gaussian　Calibration　Model(MGCM)　is　proposed　by　us,　and　a　Time　Latency　Calibration　Model(TLCM)　is　proposed　to　address　the　time　latency　effect　during　the　online　phase　for　a　fast　moving　object.　Firstly,　MGCM　is　applied　to　the　collected　RSS　data　to　precisely　extract　the　less　noised　RSS.　Then　a　feed　forward　neural　network　is　trained　to　build　a　model　between　RSS　and　physical　location.　Finally,　TLCM　is　applied　during　the　online　phase.　The　experimental　results　indicate　that　MGCM　and　TLCM　reduce　error　compared　to　traditional　positioning　method　respectively,　which　demonstrate　the　advantages　of　the　proposed　algorithms.