无线衰落信道中信道预测的误差分析  

Performance analysis of channel prediction in rapid dispersive fading channels

作　　者：汪安春[1] 佘小明 周世东[1] 许希彬[1] 姚彦[1] 

机构地区：[1]清华大学电子工程系微波与数字通信技术国家重点实验室,北京100084

出　　处：《清华大学学报（自然科学版）》2004年第10期1430-1432,1436,共4页Journal of Tsinghua University(Science and Technology)

基　　金：国家"八六三"高技术项目(2001AA123012)

摘　　要：为了分析受Gauss噪声干扰下信道预测误差方差的明确表达式,给出了采用过去无限样本时单径平衰落信道下受Gauss噪声干扰的线性预测误差的方差公式。指出了采用有限样本进行信道估计时,其信道预测误差的方差和正规方程相差一个N0,当归一化Doppler频率很高(大于0.01)时,可以采用有限个数样本(如50个),也可以得到很好的结果。给出了正交频分复用(OFDM)系统中信道线性预测误差的方差公式和具有鲁棒性的线性信道预测的方法,解决了信道线性预测的误差分析问题。Next-generation wireless communications systems will use orthogonal frequency division multiplexing (OFDM) and adaptive techniques such as dynamic resource management (DRM) and adaptive modulation coding (AMC)to improve system performance. However,the channel coefficients must be predicted in advance to realize these adaptive methods in practice. Previous papers have analyzed the prediction error, but they have not given an explicit correctformula for the variance of the prediction error. This paper gives explicit expressions for the variance of the channel prediction error both for flat-fading channels and OFDM systems, and explains how to calculate the variance of the prediction error based on a limited number of past channel samples. The analysis shows that robust channel prediction is possible.

关 键 词：正交频分复用(OFDM) 线性预测 动态资源分配(DRM) 自适应编码调制(AMC) 

分 类 号：TN911.23[电子电信—通信与信息系统]