Cyclostationarity-based signal detection and clasiffication using hidden Markov models

Detection and classification of low SNR signals are crucial for an efficient adaption of the cognitive radio to the changes in the radio environment. This paper describes threshold-based signal detection and HMM-based classification using cyclic frequency domain profile of spectral coherence. The performance of the considered radio signal detection and modulation classification, obtained using software simulation models, is shown, as well.