In　order　to　prevent　a　malicious　terminal　of　a　group　identifying　itself　and　group　fraud　to　others,　in　this　paper,　a　TPM　identification　scheme　based　on　threshold　policy　is　presented.　Its　security　is　based　on　computational　DH　problem　(CDHP)　and　K-CAA　problem,　which　can　effectively　prevent　a　malicious　terminal　of　a　group　identifying　itself　and　group　fraud　to　others.　The　scheme　also　fully　takes　into　account　the　limited　computing　power　of　TPM　to　enhance　its　computing　efficiency.