Based　on　lots　of　research　and　analysis　on　indoor　radio　signal　propagation　features　and　the　traditional　indoor　location　algorithms,　a　new　method　that　uses　BP(Back　Propagation)　neural　network　to　fit　the　indoor　radio　signal　propagation　model　is　proposed,　which　avoids　inaccurately　estimating　the　parameters　A　and　n　in　the　indoor　radio　signal　propagation　model.　Distance　value　proportional　to　the　RSSI(Received　Signal　Strength　Indicator)　input　through　the　well-trained　BP　neural　network　is　obtained,　and　then　Taylor　series　expansion　algorithm　is　used　to　determine　the　coordinates　of　the　blind　node.　Finally,　the　simulation　and　experiment　results　on　the　ZigBee　platform　verify　the　feasibility　and　effectiveness　of　the　proposed　algorithm.