In　this　paper,　we　propose　a　method　to　improve　the　performance　of　the　bias-compensated　normalized　least　mean　square　(BCNLMS)　algorithm　by　the　convex　combination　scheme.　Although　the　BCNLMS　algorithm　performs　effective　performance　under　input　noisy　case,　but　it　still　suffers　from　the　contradiction　between　the　steady-state　accuracy　and　the　convergence　speed.　To　address　this　problem,　we　introduce　the　convex　combination　scheme　into　the　BCNLMS　to　improve　the　performance　by　exploiting　advantages　of　parallel　BCNLMS　algorithm　with　different　parameters.　Our　simulations　in　different　scenarios　show　the　efficiency　of　the　proposed　convex　combination　of　the　BCNLMS　algorithm　that　can　outperforms　the　original　NLMS　and　BCNLMS.