Based on neural network spectrum prediction of cognitive radio

Cognitive radio technology enables unauthorized users to use the same spectrum in the absence of interference. And spectrum sensing for non-authorized users perceive the availability of the channel is a very important tool. But the perceived need to consume a large spectrum of energy, and this part of the energy can be reduced by perceive of the spectrum holes. The use of reliable prediction methods, non-authorized users will perceive only the channel idle, which will not only reduce energy consumption, but also spectrum efficiency can be increased. In this paper, we design a neural network prediction model of the spectrum, by simulation, spectrum occupancy state can be predicted.