Chaos and Support Vector Machines Optimization Based Wavelet Transform Weighted Multi-modulus Blind Equalization Algorithm
Fang Xu · Jisuanji fangzhen · 2013
In order to improve the blind equalization performance for high-order QAM signals,chaos and support vector machines optimization based wavelet transform weighted multiple modulus blind equalization algorithm was proposed.To avoid local minima,the proposed algorithm utilized support vector machines to initialize the weight vectors of equalizer.During the process,the parameters of support vector machines were optimized by chaos optimization algorithm.The equalizer's input signals were pretreated by using orthogonal wavelet transform to reduce the self-correlation.The proposed algorithm adapted the Weighted Multiple Modulus Algorithm(WMMA) to adjust the modulus of weight vectors.The simulation results with the underwater acoustic channel,compared with WMMA and WT-WMMA(Wavelet Transform based WMMA),show that the proposed algorithm has higher convergence speed and smaller residual mean square error.