Wireless networks communicate with each other using radio spectrum bands which are assigned to license owners. Due to the fixed spectrum assignment policy, a large portion of the spectrum stays unused. The aim of cognitive radio is enabling users which do not hold a license to be able to access the spectrum assigned to license owners. In the channel assignment problem, the objective is to assign channels to unlicensed users in order to maximize channel utilization without causing any interference to licensed users. In this study, we propose a hyper-heuristic approach to solve the channel assignment problem in cognitive radio networks. Results show that our approach gives high channel utilization rates by allowing unlicensed users to access the channels owned by licensed users. The results are promising and promote further study.