Abstract: Contraposed that the high power amplifier produced nonlinear distortion in satellite channel, this paper proposed a particle filtering blind equalization method. Advantage of this method based on particle filtering is not to linearize the nonlinear channel, but using numbers of discrete random sample points with weighted values to approximate the expected distribution. Through modeling nonlinear model as a state space model, particle filtering can track the channel parameters and estimate the emitted symbol sequence .The simulation results show that the algorithm accomplished particle filtering estimation of the amplifier nonlinear characteristic of amplitude and phase, the restoration of the symbol sequence. Compared with truncate Volterra equalization , particle filtering algorithm obtained about 1.5 dB performance gain at the bit error rate of ； it can improve the performance of the algorithm by increasing the particle number and smooth length to a certain extent; but the complexity and performance tradeoffs to consider, it can’t infinitely increase the particle number.