[AbstractC] luster structure in cognitive radio networks facilitates cooper- ative spectrum sensing, routing and other functionalities. The unlicensed channels, which are available for every member of a group of cognitive radio users, consolidate the group into a cluster, and the availability of unlicensed channels decides the robustness of that cluster against the licensed users’ influence. This paper analyses the problem that how to form robust clusters in cognitive radio network, so that more cognitive radio users can get benefits from cluster structure even when the primary users’ operation are intense. We provide a formal description of robust clustering problem, prove it to be NP-hard and propose a centralized solution, besides, a distributed solution is proposed to suit the dynamics in the ad hoc cognitive radio network. Congestion game model is adopted to analyse the process of cluster formation, which not only contributes designing the distributed clustering scheme directly, but also provides the guarantee of convergence into Nash Equilibrium and convergence speed. Our proposed clustering solution is versatile to fulfill some other requirements such as faster conver- gence and cluster size control. The proposed distributed clustering scheme outperforms the related work in terms of cluster robustness, convergence speed and overhead. The extensive simulation supports our claims.