S波段两状态LMS信道模型的自适应长期预测

An adaptive long-range prediction based on two-state LMS channel model at S-band

针对S波段模型参数可变的窄带两状态陆地移动卫星信道模型,基于加权预测思想提出一种自适应长期预测方法.首先将卫星通信下行链路的阴影遮蔽建模为两状态马尔科夫链的Gilbert-Elliot信道模型,然后利用加权预测思想预测未来长期内的信道状态,并基于最小均方算法由迭代自适应跟踪方法更新线性自回归模型的系数,进而预测出未来的信道衰落序列.研究结果表明:该方法能精确地预测出未来长期内的信道状态和衰落序列,且相比长期预测方法,改善预测性能,并具有实时性和低复杂度优点,可用于窄带LMS通信系统自适应传输性能分析.

Considering the narrowband two-state land mobile satellite channel model with variable model parameter at S-Band, an adaptive long-range prediction method is proposed based on weighting prediction. Firstly, a two-state Markov Gilbert-Elliot channel model with an ability of describing shadowing conditions of satellite communication downlink is established. And then, the future long-range channel state is predicted by weighting prediction, and the coefficients of linear auto-regression model are updated by iterative adaptive tracking method using the least mean square algorithm. Finally, the future channel fading series are predicted. Simulation results show that the proposed method not only can be used to predict the future long-range channel states and fading series accurately, but also improve prediction performance compared with the long-range prediction method. Moreover, this method has ability of real-time and low-complexity and can be used in the adaptive transmission performance analysis of narrowband LMS communication systems.