Abstract:

Under the time-varying fading channel, an adaptive modulation recognition algorithm is presented. A new dynamic state space model is designed to describe timevarying characteristics of modulation schemes and channel gain. A first-order finite state Markov channel （FSMC） model is introduced for the fading channel. On this basis, a new algorithm, which adopts the sequential Bayesian inference method and is proposed to fully exploit the dynamic transfer characteristics of the hidden channel state, achieves joint estimation of modulation and time-varying channel gain. The simulation results prove that performance of the algorithm compared to traditional ALRT algorithms greatly improves, and increasing the number of sampling points or reducing the Doppler shift value can make the performance better.