Research on the channel prediction algorithm of multiwavelet combined with support vector machines based on FPGA

By applying the chaotic theory and reconstructing embedding phase space, utilizing support vector machine method to establish prediction model for fading channel signal, and using multi-wavelet theory to support vector machine algorithm, then proposing that multi-wavelet support vector machine method should apply to the fading channel signal prediction. Based on investigating those four problems: (1)kernel function selection problem, (2)algorithm of IP core implementation problem, (3)expansion of large-scale auto-correlation matrix equation for solving by FPGA problem, (4)FPGA-based embedding dimension and iterative algorithm of time delay problems, we can verify their effectiveness by designing module verification, simulation and physical verification platform. It provide authentication and technical support for the development of new communication systems and application of new technologies.