ANN Based Spectrum Sensing Technique for Cognitive Radio Applications

An ANN based spectrum sensing technique for audio FM and the wireless microphone signals in TV band is proposed. The artificial neural network (ANN) model trained with the autocorrelation peaks of the signal in channel identifies it as a white space or a primary signal. The performance of this technique is efficient in terms of false alarm rate and probability of detection; the proposed method presents less mathematical complexity as compared to other recent spectrum sensing techniques. Simulation results are presented.