基于多簇模型的自适应OFDM系统信道预测方法  

Adaptive OFDM channel predictor based on multi-cluster model

作　　者：张焱[1] 胡昕炜[1] 肖立民[2] 周世东[1] 王京[2] 

机构地区：[1]清华大学电子工程系,北京100084 [2]清华大学信息技术研究院,北京100084

出　　处：《电波科学学报》2010年第6期1027-1033,共7页Chinese Journal of Radio Science

基　　金：国家科技重大专项课题(2009ZX03007-003);国家重点基础研究发展规划项目资助(2007CB310608);国家863项目资助(2009AA011501)

摘　　要：基于多簇宽带无线信道模型,提出了一种用于自适应正交频分复用(OFDM)系统的信道预测方法。利用无线传播环境中多径成簇到达的特征,将信道预测过程分成学习、估计和预测三个阶段,在通信开始阶段,通过发送大量导频得到簇的个数和时延参数,在通信过程中,根据簇参数对导频个数和图案进行自适应调整,结合自回归方法,预测未来信道变化。本方法可以降低导频使用量,提高系统频谱利用效率,其预测性能通过蒙特卡洛仿真得到了验证。A channel predictor based on multi-cluster model was presented for adaptive orthogonal frequency division multiplexing（OFDM）system.This predictor utilized the characteristic that several paths usually arrive in the same cluster in wireless propagation environment.The channel prediction process was divided into three stages including learning,estimation and prediction.Firstly the number and the delays of the clusters were estimated by sending plentiful pilots before communication.Then during the communication process,the number and the pattern of pilots can be adaptively adjusted.Combined with the auto-regress method,the future channel state information can be predicted.This predictor can reduce the number of pilots used and improve system spectral efficiency.This new predictor＇s performance has been proved by Monte-Carlo simulation results.

关 键 词：正交频分复用 信道预测 多簇 自适应预测 

分 类 号：TN929.5[电子电信—通信与信息系统]