A Method of Dynamic Spectrum Access in HD Radio Based on BP Neural Network

The dynamic spectrum access method based on human ear perception can keep the quality of analog signal in HD Radio (Hybrid Digital Radio) stable, but it takes too much time to meet the real-time requirement of Radio. This paper proposes a dynamic spectrum access method in HD Radio based on BP NN (Back Propagation Neural Network). In the method, the data of dynamic spectrum access method based on human ear perception are used to the train BP NN, according to audio characteristic values, positions of dynamic digital spectrum can be got by the trained BP NN. The simulation results show that the analog signals quality of human ear perceived in HD Radio can improved by the method proposed, and the processing speed can meet the real-time requirement of Radio.