In　this　paper,　by　employing　primary/backup　(P/B)　technique　and　the　concept　of　game　theory,　we　propose　a　game　theoretical　fault-tolerant　task　allocation　algorithm　for　heterogeneous　WSNs.　During　the　experiment,　our　job　can　be　divided　into　two　parts.　Firstly,　our　algorithm　introduces　the　idea　of　game　theory　into　the　solving　of　primary　tasks　allocation　problem,　then　we　define　the　playoffs　function　to　find　the　Nash　equilibrium　point　of　task　allocation　for　the　purpose　of　saving　the　node　energy　consumption　and　balancing　the　network　energy　under　the　restraint　of　tasks　deadline.　Secondly,　based　on　the　allocation　of　primary　tasks,　our　algorithm　determinate　the　scheme　of　backup　copies,　then　allocate　those　tasks　by　employing　passive　backup　copies　overlapping　technology.　The　result　of　simulations　shows　the　feasibility　and　effectiveness　of　our　algorithm.