Constellation shape as a robust signature for digital modulation recognition

This work reports on a new digital modulation recognition algorithm in an AWGN channel and in the presence of carrier recovery errors. The proposed classification technique uses the signal constellation shape as a stable modulation signature. The recovered constellations are modeled by binomial nonhomogenous spatial random fields and are used in an ML classifier. We experimentally show that shape as a global signature exhibits considerable stability in high noise, weakly synchronous environments. Experimental results are shown for various modulation standards including V.29, V.29 fallback, 8-PSK and 16-QAM.