Cognitive Wi-Fi 2.0 wireless networks with the typical characteristics of cognition, reconfirmation and self-organizing abilities as the novel open spectrum sharing technology, that are used to improve the network performance, have promisingly captured a great attention from the wireless community. In this paper, to capture the hierarchical decision-making properties of the power control, the study proposes a dynamic hierarchical power control for the cognitive Wi-Fi 2.0 wireless networks. The spectrum sharing issue is formulated as the Stackelberg game with the more foresighted cognitive users as the leaders and short-sighted users as follower, and the paper derives the closed-form power policy, respectively. Meanwhile, the study concludes that this policy can achieve the optimal performance in the weak interference environment after analysis. Simulations results show that the proposed algorithm converges after limited iterations, which also effectively guarantees the optimal trade-off between the individual performance and the network overall performance.