Spectrum anomalies autonomous detection in cognitive radio using Hidden Markov Models

The precisely detection of electromagnetic spectrum anomaly is important and crucial for increasing demand on spectrum security, especially in the condition of complex electromagnetic environment and lack of pre-knowledge information about frequency use. There were many research methods of anomalies detection to conquer malicious radio events. In this paper, we proposed a spectrum anomalies autonomous detection and classification method based on spectrum amplitude probability and Hidden Markov Model (HMM) to cover the shortage of passive spectrum anomaly detection on site at present. We trained and tested the method through experiments using real spectrum measurement data. The experimental results show that new approach performs well for recognizing different kinds of spectrum anomalies with rather high accuracy.