A hierarchical approach to the classification of digital modulation types in multipath environments

Abstract

This study presents a hierarchical approach to the classification of digital modulation schemes of types [2,4,8]-PSK, [2,4,8]-FSK and [16,64,256]-QAM in various SNR levels and multipath propagation channel conditions. A tree-based classification approach is selected as it leads to a relatively simple overall scheme with a few higher-order statistics parameters selected as discriminating features. Back propagation neural network decision units are adopted at each tree node to offer the flexibility needed to cope with varying propagation environments for all schemes considered, except for specific MQAM types, which are discriminated via equalization algorithms. Extensive simulations are conducted in various multipath environments, and show overall classification performances to be strongly affected by the amount of multipath distortion and noise in the transmission channels.