Although　5G　and　beyond　communication　technology　empower　a　large　number　of　edge　heterogeneous　devices　and　applications,　the　stringent　security　remains　a　major　concern　when　dealing　with　the　millions　of　edge　computing　tasks　in　the　highly　dynamic　heterogeneous　networks　(HDHNs).　Blockchains　contribute　significantly　to　addressing　security　challenges　by　guaranteeing　the　reliability　of　data　and　information.　Since　the　node＇s　mobility,　there　are　risks　of　exiting　the　network　and　leaving　the　remaining　tasks　noncomputed.　Therefore,　we　model　the　cost　function　of　offloaded　computing　tasks　as　a　dynamic　stochastic　game.　To　reduce　the　computational　complexity,　the　Time-Variant　Mean-Field　term　(TVMF)　is　adopted　to　solve　the　cost-optimized　problem.　What＇s　more,　we　design　an　Adaptivity-Aware　Practical　byzantine　fault　tolerance　consensus　Protocol　(AAPP)　to　dynamically　formulate　domains,　execute　leader　node　selection　with　regard　to　task　completion　and　quickly　verify　computational　results.　In　addition,　a　Dynamic　Multi-domain　Fractional　Repetition　uncoded　repair　storage　(DMFR)　scheme　with　variant　redundancy　is　proposed　to　reduce　the　storage　pressure　and　repair　overhead.　The　simulation　is　implemented　to　demonstrate　our　scheme　outperforms　the　benchmarks　in　terms　of　cost　and　time　overhead.