Fog　computing　is　an　encouraging　technology　in　the　coming　generation　to　pipeline　the　breach　between　cloud　data　centers　and　Internet　of　Things　(IoT)　devices.　Fog　computing　is　not　a　counterfeit　for　cloud　computing　but　a　persuasive　counterpart.　It　also　accredits　by　utilizing　the　edge　of　the　network　while　still　rendering　the　possibility　to　interact　with　the　cloud.　Nevertheless,　the　features　of　fog　computing　are　encountering　novel　security　challenges.　The　security　of　end　users　and/or　fog　nodes　brings　a　major　dilemma　in　the　implementation　of　real　life　scenario.　Although　there　are　several　works　investigated　in　the　security　challenges,　physical　layer　security　(PLS)　in　fog　computing　is　not　investigated　in　the　above.　The　distinctive　and　evolving　IoT　applications　necessitate　new　security　regulations,　models,　and　evaluations　disseminated　at　the　network　edge.　Notwithstanding,　the　achievement　of　the　current　cryptographic　solutions　in　the　customary　way,　many　aspects,　i.e.,　system　imperfections,　hacking　skills,　and　augmented　attack,　has　upheld　the　inexorableness　of　the　detection　techniques.　Hence,　we　investigate　PLS　that　exploits　the　properties　of　channel　between　end　user　and　fog　node　to　detect　the　impersonation　attack　in　fog　computing　network.　Moreover,　it　is　also　challenging　to　achieve　the　accurate　channel　constraints　between　end　user　and　fog　node.　Therefore,　we　propose　Q-learning　algorithm　to　attain　the　optimum　value　of　test　threshold　in　the　impersonation　attack.　The　performance　of　the　propose　scheme　validates　and　guarantees　to　detect　the　impersonation　attack　accurately　in　fog　computing　networks.