APP下载

基于本地人工信道的新型OFDM信道估计方法

2014-09-17张家辉解永生李宝清

现代电子技术 2014年17期
关键词:奇异值分解

张家辉 解永生 李宝清

摘 要: 介绍了一种新颖的基于本地人工信道的信道估计方法,并对该方法进行进一步的奇异值分解和简化,使其具有较少的存储量和计算量,具有很大的实际应用价值。对相关方法进行了仿真和比较,在误比特率性能上与传统线性最小均方误差估计相差小于1 dB。

关键词: OFDM; 信道估计; 本地人工信道; 奇异值分解

中图分类号: TN92?34 文献标识码: A 文章编号: 1004?373X(2014)17?0013?03

Abstract: A novel channel estimation method based on local artificial channel is introduced in this paper. The singular value decomposition and simplification of the method were conducted to make it have small storage content, calculated amount and high application value. The relevant methods were simulated and compared with the novel one. The results shown that its bit error rate is only 1 dB less than that of the traditional linear minimum mean?squared error estimation method.

Keyword: OFDM; channel estimation; local artificial channel; singular value decomposition

0 引 言

正交频分复用(OFDM)技术由于具有较高的频谱利用率、对抗多径干扰效果好、较高的数据速率等优点,广泛地应用到无线通信系统中。因为无线信道具有多径和时变特性,为了降低多径和衰弱对系统性能的影响,需要在接收端利用信道估计来补偿信道的变化。

OFDM信道估计是一个古老的话题,但是对于它的研究一直没有停止,有对估计算法的研究,如文献[1]提出的基于本地人工信道的估计方法;还有设计合适导频序列来对抗频率偏移的信道估计[2];有针对MIMO系统的递归更新的信道估计方法[3];有对导频摆放方法的研究[4];有基于二维扩展的对时域快变信道的估计[5];当然也有对现有算法的改进的文章[6?7]。本文主要针对基于导频的信道估计算法的研究,这类算法是利用在数据流中插入一定数量的已知数据(导频)进行信道估计,传统估计方法包括最小二乘估计(LS)和线性最小均方误差估计(LMMSE),两种方法在性能和复杂度上各有利弊。本文介绍了这两种基于导频的OFDM信道估计基本方法,以降低LMMSE复杂度为目标,引入一种基于本地人工信道的估计方法来减少LMMSE方法中的自相关矩阵问题,并对此改进方法进行进一步的奇异值分解解决矩阵求逆的问题,使本文的方法不但具有很少的计算量而且在性能上与传统LMMSE方法相差不到1 dB,是一种具有实际应用价值的OFDM信道估计方法。

需要在接收端存储人工信道[G(N×1)、]部分的奇异值([P]个)和奇异向量组成的矩阵[(N×P),]并在接收端增加一个估计SNR的模块。这样就把普通的LMMSE简化成为只需要进行简单的相乘相加的运算,而舍去了自相关矩阵的计算和复杂的矩阵求逆运算,而这仅仅需要增加一小部分的数据存储。

3 仿 真

用Matlab进行仿真,接收端只考虑信道估计的性能,不计其他纠错编码的影响。子载波数[N=]1 024,CP长度为[N4]256,采样率为10 MHz,采用块状导频结构。实际信道采用COST 207标准的典型城区Ⅰ多径信道[10],多普勒频移为25 Hz,本地人工信道多径延时满足均匀分布,功率延时谱满足负指数分布,应用与中低速的图传系统。

图4显示了几种AC(Artificial Channel)方法的性能,其中Delay Sufficient表示的是按照文中的要求构造本地人工信道,即本地人工信道的[τmax]设为CP的长度(远大于实际信道);而Delay Insufficient1和Delay Insufficient2都与文中的要求有一定的差距,情况1的本地人工信道的[τmax]略小于实际信道的[τmax,]而情况2的本地人工信道的[τmax]为实际信道的[τmax]的一半。可以看出,当本地人工信道设置合理时,它的性能基本跟实际的LMMSE算法接近,相同BER信噪比差距小于1 dB;而当本地人工信道设置不合理时,性能会有很大的差距,如情况1在10 dB之后开始逐渐地变差,16 dB之后甚至比LS算法还要差;而情况2由于人工信道设置得更不合理,导致从8 dB开始就比LS算法差了。

图5显示AC和SVD结合后的性能,其中AC SVD1表示的是按文中要求进行取舍奇异值,即取了最大的前CP个奇异值;而AC SVD2只取了最大的前CP/2奇异值。从图中可以看出,SVD1与普通的AC基本重合,可以认为基本没有性能上的损失,而SVD2因为少取了一些奇异值,丢弃了一些有用信息而导致性能下降,尤其是在信噪比10 dB之后性能迅速下降,16 dB之后甚至比LS算法的性能还要差。

4 结 论

通过以上仿真可以看出,本文提出的AC+SVD的方法不但减少了求自相关矩阵、矩阵求逆在内的大量计算,只需要进行简单的乘法和加法,而且在性能上跟传统的LMMSE算法差距在1 dB以内,所以该方法是一种具有实际应用价值的、高性能的OFDM信道估计方法。

参考文献

[1] SAVAUX Vincent, SKRZYPCZAK Alexandre, LOUET Yves, et al. Near LMMSE channel estimation preformance with artificial channel at receiver for OFDM systems [C]// 13th International Workshop on Signal Processing in Wireless Communications. [S.l.]: [s.n.], 2012: 545?549.

[2] OLIVER J, ARAVIND R, PRABHU K M M. Improved least squares channel estimation for orthogonal frequency division multiplexing [J]. IET Signal Processing, 2012, l.6: 45?53.

[3] HESKETH Thomas, DE LAMARE Rodrigo C, WALES Stephen. Adaptive MMSE channel estimation algorithms for MIMO system [EB/OL]. [2012?11?26]. www.ymcn.org/d?9bs1.html.

[4] Qun Yu and Ronglin Li.”Research on Pilot Pattern Design of Channel Estimation”Journal of Automation and Control Engineering,Vol.1,No.2,March 2013.

[5] PENA?CAMPOS F, CARRASCO?ALVAREZ R, LONGORIA?GANDARA O, et al. Estimation of fast time?varying channels in OFDM systems using two?dimensional prolate [J]. IEEE Transaction on wireless communications, 2013, 12(2): 898?907.

[6] ZHOU Wen, LAM Wong Hing. A fast LMMSE channel estimation method for OFDM systems [D]. Hong Kong, China: Department of Electrical and Electronics Engineering, The University of Hong Kong, 2009.

[7] MINN H, BHARGAVA V K. An investigation into time?domain approach for OFDM channel estimation [J]. IEEE Transactions on Broadcasting, 2009, 46(4): 240?248.

[8] EDFORDS O, SANDELL M, VAN DE BEEK J J, et al. OFDM channel estimation by singular value decomposition [J]. IEEE Transactions on Communications, 1998, 46(7): 923?927.

[9] VAN DE BEEK J J, EDFORS O, SANDELL M, et al. On Channel Estimation in OFDM Systems [C]// Proceedings of IEEE conference on Vehicular Technology. Chicago, USA: IEEE, 1995, 2: 815?819.

[10] 杨大成.移动传播环境[M].北京:机械工业出版社,2003.

参考文献

[1] SAVAUX Vincent, SKRZYPCZAK Alexandre, LOUET Yves, et al. Near LMMSE channel estimation preformance with artificial channel at receiver for OFDM systems [C]// 13th International Workshop on Signal Processing in Wireless Communications. [S.l.]: [s.n.], 2012: 545?549.

[2] OLIVER J, ARAVIND R, PRABHU K M M. Improved least squares channel estimation for orthogonal frequency division multiplexing [J]. IET Signal Processing, 2012, l.6: 45?53.

[3] HESKETH Thomas, DE LAMARE Rodrigo C, WALES Stephen. Adaptive MMSE channel estimation algorithms for MIMO system [EB/OL]. [2012?11?26]. www.ymcn.org/d?9bs1.html.

[4] Qun Yu and Ronglin Li.”Research on Pilot Pattern Design of Channel Estimation”Journal of Automation and Control Engineering,Vol.1,No.2,March 2013.

[5] PENA?CAMPOS F, CARRASCO?ALVAREZ R, LONGORIA?GANDARA O, et al. Estimation of fast time?varying channels in OFDM systems using two?dimensional prolate [J]. IEEE Transaction on wireless communications, 2013, 12(2): 898?907.

[6] ZHOU Wen, LAM Wong Hing. A fast LMMSE channel estimation method for OFDM systems [D]. Hong Kong, China: Department of Electrical and Electronics Engineering, The University of Hong Kong, 2009.

[7] MINN H, BHARGAVA V K. An investigation into time?domain approach for OFDM channel estimation [J]. IEEE Transactions on Broadcasting, 2009, 46(4): 240?248.

[8] EDFORDS O, SANDELL M, VAN DE BEEK J J, et al. OFDM channel estimation by singular value decomposition [J]. IEEE Transactions on Communications, 1998, 46(7): 923?927.

[9] VAN DE BEEK J J, EDFORS O, SANDELL M, et al. On Channel Estimation in OFDM Systems [C]// Proceedings of IEEE conference on Vehicular Technology. Chicago, USA: IEEE, 1995, 2: 815?819.

[10] 杨大成.移动传播环境[M].北京:机械工业出版社,2003.

参考文献

[1] SAVAUX Vincent, SKRZYPCZAK Alexandre, LOUET Yves, et al. Near LMMSE channel estimation preformance with artificial channel at receiver for OFDM systems [C]// 13th International Workshop on Signal Processing in Wireless Communications. [S.l.]: [s.n.], 2012: 545?549.

[2] OLIVER J, ARAVIND R, PRABHU K M M. Improved least squares channel estimation for orthogonal frequency division multiplexing [J]. IET Signal Processing, 2012, l.6: 45?53.

[3] HESKETH Thomas, DE LAMARE Rodrigo C, WALES Stephen. Adaptive MMSE channel estimation algorithms for MIMO system [EB/OL]. [2012?11?26]. www.ymcn.org/d?9bs1.html.

[4] Qun Yu and Ronglin Li.”Research on Pilot Pattern Design of Channel Estimation”Journal of Automation and Control Engineering,Vol.1,No.2,March 2013.

[5] PENA?CAMPOS F, CARRASCO?ALVAREZ R, LONGORIA?GANDARA O, et al. Estimation of fast time?varying channels in OFDM systems using two?dimensional prolate [J]. IEEE Transaction on wireless communications, 2013, 12(2): 898?907.

[6] ZHOU Wen, LAM Wong Hing. A fast LMMSE channel estimation method for OFDM systems [D]. Hong Kong, China: Department of Electrical and Electronics Engineering, The University of Hong Kong, 2009.

[7] MINN H, BHARGAVA V K. An investigation into time?domain approach for OFDM channel estimation [J]. IEEE Transactions on Broadcasting, 2009, 46(4): 240?248.

[8] EDFORDS O, SANDELL M, VAN DE BEEK J J, et al. OFDM channel estimation by singular value decomposition [J]. IEEE Transactions on Communications, 1998, 46(7): 923?927.

[9] VAN DE BEEK J J, EDFORS O, SANDELL M, et al. On Channel Estimation in OFDM Systems [C]// Proceedings of IEEE conference on Vehicular Technology. Chicago, USA: IEEE, 1995, 2: 815?819.

[10] 杨大成.移动传播环境[M].北京:机械工业出版社,2003.

猜你喜欢

奇异值分解
基于SVD分解的主成分分析人脸识别方法
基于奇异值分解的银行客户数据隐私保护算法研究
k—means聚类算法在提高图书馆数字文献服务效能中的应用
结合PCA及字典学习的高光谱图像自适应去噪方法
基于分块DWT和SVD的鲁棒性数字水印算法
一种基于奇异值分解的鲁棒水印算法
基于HOG—SVD特征的人脸识别
基于奇异熵和随机森林的人脸识别
基于SVD确定NMF初始化矩阵维数
协同过滤算法改进及研究