Nonlinear Prediction of Fast Fading Channel Based on Minimax Probability Machine

Adaptive modulation technique has been widely used in wireless communication systems and channel prediction plays an important role in adaptive modulation technique. Minimax probability machine shows good performance in classification and prediction by controlling the generalization error boundary and trying to make it lowest. In this paper, we introduce a nonlinear prediction algorithm of fast fading channel based on minimax probability machine (MPM). In this algorithm, the learning samples are constructed on the observed values of Jakes fading channel coefficients. After selecting the best embedding dimension of learning samples, it obtains predictive values of time series by the nonlinear prediction which is implemented by the minimax probability machine. The simulation result shows that the proposed algorithm can make the accuracy of estimation and has good real-time performance. Compared with the support vector machine (SVM) , MPM has more accurate prediction and faster training speed.