Comprehensive features based digital modulation identification using a neural tree network

A neural tree network (NTN) based automatic digital modulation identification method is demonstrated for 13 types of digital modulation using comprehensive features, including power spectral features, cyclic spectral features and high-order cumulant features. The NTN is a self-organizing, hierarchical classifier implementing a sequential linear strategy and requiring no statistical analysis of the features. Experiments show that these modulations can be recognized at a SNR of 5 dB in AWGN, and this method also works well for frequency modulations and some amplitude-phase modulations in multipath environments.