In order to realize a variety of intelligent applications, a large number of devices need to collect and transmit potentially large amounts of data through the Internet of Things. However, with the explosive growth of communication devices, spectrum resources are becoming more and more scarce, which in turn makes it impossible to send information in time. Cognitive radio technology is an effective way to resolve spectrum allocation shortages, while the spatiotemporal uncertainty of spectrum has a great influence on the secondary user (SU)'s throughput. In this paper, we investigate the network coding scheme in cognitive radio networks under multiple primary and SU relationships, including non-collaboration, collaboration, and lease. Moreover, we propose a hybrid network coding scheme to improve the utilization efficiency of idle spectrum. In particular, the successful transmission rate and the packet loss rate can be increased and decreased through redundant coding, respectively. The simulation results show that our proposed scheme can achieve a better performance in terms of both delay and throughput than the existing mechanisms.