支持向量机在衰落信道预测中的应用

研究信道通信优化控制问题,针对传统衰落信道预测算法中的不足,提出一种改进LSSVM的衰落信道预测。首先通过相空间重构对衰落信道系数序列进行重构,然后采用LSSVM对训练样本集进行学习,并通过自适应遗传算法对LSSVM参数进行优化建立最优衰落信道系数预测模型,最后采用测试集对模型的性能进行验证。仿真结果表明,相对于传统LSS-VM模型,改进模型提高了衰落信道系数预测精度,是一种进行衰落信道非线性预测的有效方法。

Application of Prediction of Fading Channels Based on Least Square Support Vector Machine

Because Least Squares Support Vector Machine(LSSVM) has some shortages in the prediction of fading channels,this paper put forward a fading channel prediction method based on improved LSSVM.Firstly,fading channel coefficients wre reconstructed by the phase space reconstruction.Then raining sample set was input into LSSVM to learn while LSSVM parameters were optimized by adaptive genetic algorithm to establish the optimal prediction model for channel fading coefficient.Finally,the model was test by the validation test set.The simulation results show that,compared with the traditional LSSVM model,the proposed model improves the channel fading coefficient prediction precision,and is an effective method for the fading channel prediction.