In this paper, we propose two schedulers for channel sensing in cognitive radio networks. We use both of rules‐based learning and statistical‐based learning to develop the proposed solutions. Both of these schedulers take three channel parameters as an input to generate scheduling decision. These parameters are utilization, success rate, and channel quality. First proposed scheduler employs fuzzy inference technique to generate the decision regarding channel sensing order. It makes use of genetic algorithm to find the best set of fuzzy rules based on environment dynamics. The second proposed scheduler uses Bayesian inference concept by applying Baye's rule on the perceived distribution function of channel occupancy to improve its accuracy. To track the advancement of scheduler performance, we devised a measurement concept called convergence indicator. Both of the proposed mechanisms showed remarkable performance and adaptability to the highly dynamic behavior of wireless environment. Fuzzy inference solution achieved on average 138 per cent increases in terms of channel utilization compared to fixed sensing order, while Bayesian inference achieved 85 per cent increase. Copyright © 2014 John Wiley & Sons, Ltd.