This　paper　presents　a　new　target　tracking　and　the　possible　changing　bias　registration　approach　based　on　sensor　selection　for　large-scale　distributed　sensor　networks.　We　try　to　address　this　target　tracking　problem　at　the　following　three　steps.　Firstly,　local-level　tracking　is　addressed　based　on　the　estimated　sensor　biases　for　each　sensor,　and　only　the　state　estimates　are　transmitted　to　the　fusion　center;　Secondly,　data　fusion　is　carried　out　by　using　the　sensor　selection　approach　at　the　fusion　center,　target　state　is　estimated　based　on　the　tracking　results　reported　by　the　selected　sensors;　Finally,　sensors＇　biases　are　updated　at　the　fusion　center.　In　addition,　both　of　sensor　coverage　problem　and　the　possible　changing　bias　problem　are　considered　in　our　paper.　The　proposed　approach　only　needs　to　select　a　small　number　of　sensors　for　tracking,　rather　than　the　traditional　approaches,　which　prefer　to　use　all　of　the　sensors.　Simulation　results　show　the　effectiveness　of　the　proposed　approach.