For　the　occurrence　of　network　attacks,　the　most　important　thing　for　network　security　managers　is　how　to　conduct　attack　security　defenses　under　low-risk　control.　And　in　the　attack　risk　control,　the　first　and　most　important　step　is　to　choose　the　defense　node　of　risk　control.　In　this　paper,　aiming　to　solve　the　problem　of　network　attack　security　risk　control　under　complex　networks,　we　propose　a　game　attack　risk　control　node　selection　method　based　on　game　theory.　The　method　utilizes　the　relationship　between　the　vulnerabilities　and　analyzes　the　vulnerability　intent　information　of　the　complex　network　to　construct　an　attack　risk　diffusion　network.　In　order　to　truly　reflect　the　different　meanings　of　each　node　in　the　attack　risk　diffusion　network　for　attack　and　defense,　this　paper　uses　the　host　vulnerability　attack　and　defense　income　evaluation　calculation　to　give　each　node　in　the　network　its　offensive　and　defensive　income.　According　to　the　above-mentioned　attack　risk　spread　network　of　offensive　and　defensive　gains,　this　paper　combines　game　theory　and　maximum　benefit　ideas　to　select　the　best　Top　defense　node　information.　In　this　paper,　The　method　proposed　in　this　paper　can　be　used　to　select　network　security　risk　control　nodes　on　complex　networks,　which　can　help　network　security　managers　to　play　a　good　auxiliary　role　in　cyber　attack　defense.　©　2018　Association　for　Computing　Machinery.