摘要
提出了一种基于支持向量机的衰落信道预测算法.在相空间中构建学习样本,然后借助支持向量机的学习与判决能力实施预测.对Jakes衰落信道的预测实验表明,该算法是有效的.同时也表明嵌入维数对预测准确度有着较大影响.
The SVM is used to resolve the problem of fading channel prediction in this paper, thus a prediction algorithm for fading channels based on SVM is proposed. In our proposed algorithm, the learning samples are constructed in the phase space, and the prediction is implemented by resorting to the learning ability of the SVM. Performance evaluation of the proposed algorithm is carried out on Jakes fading channels. The results demonstrate the efficiency of the algorithm. In addition, the experiments illustrates that the embedding dimension has heavy influences on the prediction accuracy.
出处
《微电子学与计算机》
CSCD
北大核心
2009年第7期180-182,共3页
Microelectronics & Computer
基金
西安科技计划项目(CXY08012-1)
关键词
相空间
支持向量机
Jakes信道
预测
phase spaces
support vector machines (SVM)
Jakes channels
prediction