Sequential Monte Carlo-based constellation classification for ISI channels: Superconstellation method

The digital modulation classification for phase-amplitude-modulated signals transmitted through intersymbol interference (ISI) channels is considered. A classification method is proposed based on the sequential Monte Carlo (SMC) framework. In the proposed method, samples drawn from a superconstellation consisting of all the possible constellations are first evolved through the sequential importance sampling and resampling mechanism, then the constellation classification is achieved according to the distribution of the drawn samples. Moreover, digital modulation classification can be achieved along with joint estimation of transmitted data symbols and channel taps. Simulations show that the proposed classification method performs well on various constellations with different cardinalities, as well as constellations with symbols.