Research On Cognitive Radio Spectrum Prediction Based On Fuzzy Wavelet Neural Network

The proposed cognitive radio technology has improved the spectrum utilization to some extent.However,in traditional spectrum sensing,cognitive users' perception of the entire spectrum causes a lot of energy loss and processing delay.For this problem,spectral prediction techniques have received extensive attention from researchers.Spectrum prediction technology can reduce the data transmission conflict between cognitive users and primary users,avoid interference with primary user communications,reduce response delay,and increase network throughput.When using BP neural network for prediction,there is a problem of slow convergence and insufficient prediction accuracy.This paper studies the application of fuzzy wavelet neural network in spectrum prediction of cognitive radio,because the fuzzy wavelet neural network combines the good time-frequency localization properties of wavelet transform,fuzzy reasoning and neural network learning ability,it has better adaptability and prediction accuracy.The simulation results show that compared with the traditional BP neural network prediction algorithm,the prediction performance is effectively improved.Next,it analyzes the prediction performance of the algorithm under different traffic intensity,the FWNN prediction model has a smaller probability of false prediction.For cognitive radio spectrum sensing,spectrum sensing based on spectrum prediction is studied.Increasing spectrum prediction can reduce the perceived delay and reduce the interference rate.Finally,the influence of time-domain correlation on spectral prediction performance under multi-channel conditions is studied.