Multi-receiver modulation classification for satellite communications signals

In order to improve the performance of modulation classification systems, the idea of multi-receiver recognition has been developed recently. In this paper multiple receivers´ collaboration at different information levels is investigated for classification of signals used in DVB-S2 standard. Three methods are proposed for receivers´ cooperation at each one of signal, feature and decision levels. The proposed methods use cumulants and MLP neural network as signal features and classifier respectively. These methods are evaluated and compared through performance, complexity and equipment. The results show that receivers´ cooperation at signal level offers more accurate classification compared to feature and decision levels, in addition to less computational complexity.