Modulation Classification of MQAM Signals in Multipath Environment Using Mathematical Morphology

This paper proposes a new method which combines blind equalization and morphology based image-processing to identify MQAM signals in multipath environment. The advantage of our approach is that, by using image-processing, only 128×128 bits data is required to be stored. Furthermore, even the eye diagram of equalized signal for 64QAM is not opened enough, it can still be used to classify. Computer simulations are made to evaluate the performance and result shows that the new approach can reach better classification accuracy than the existing methods.