In recent years, with the development of mobile communications and Wireless Local Area Network (WLAN) as well as restrictions on limited spectrum resources, wireless spectrum resources are increasingly strained. Cognitive radio, put forward as a concept of dynamic use of spectrum, solved the problem of low spectrum utilization rate brought by the current static spectrum allocation scheme and greatly improved the utilization of the existing spectrum resources. In order to overcome the shortcomings of traditional spectrum sensing and improve spectrum detection performance under low signal-noise rate, this paper proposed a spectrum perception algorithm based on quantum neural network (QNN) and carried out an optimization study on the spectrum sensing algorithm of cognitive radio. Through the simulation experiment, we found that the improved QNN algorithm showed more excellent convergence performance and detection capability.