J4 ›› 2012, Vol. 39 ›› Issue (2): 24-28+50.doi: 10.3969/j.issn.1001-2400.2012.02.005

• 研究论文 • 上一篇    下一篇

一种低复杂度LMMSE信道估计算法

石磊1;郭宝龙1;李小平2;吕良1
  

  1. (1. 西安电子科技大学 智能控制与图像工程研究所,陕西 西安  710071;
    2. 西安电子科技大学 机电工程学院,陕西 西安  710071)
  • 收稿日期:2011-03-28 出版日期:2012-04-20 发布日期:2012-05-21
  • 通讯作者: 石磊
  • 作者简介:石磊(1981-),男,西安电子科技大学博士研究生,E-mail: shilei2002yoda@163.com.
  • 基金资助:

    国家自然科学基金资助项目(61003196,60603010)

Novel low complexity LMMSE channel estimation algorithm

SHI Lei 1;GUO Baolong 1;LI Xiaoping 2;LV Liang 1
  

  1. (1. Research Inst. of Intelligent Control & Image Engineering, Xidian Univ., Xi'an  710071, China;
    2. School of Mechano-electronic Engineering, Xidian Univ., Xi'an  710071, China)
  • Received:2011-03-28 Online:2012-04-20 Published:2012-05-21
  • Contact: SHI Lei

摘要:

针对传统线性最小均方误差(LMMSE)信道估计器矩阵求逆运算复杂度大的问题,提出了一种以相关带宽为准则提取信道自相关矩阵主要信息的子阵分块算法.根据信道相关带宽所计算的分块尺度将信道自相关矩阵分割为若干子块,包括非重叠分块法和重叠分块法.运算求逆过程中仅利用表征信道主要信息量的低频对角子阵,而忽略其他表征信道高频信息的子阵,可有效降低LMMSE算法中大的自相关信道矩阵求逆运算所带来的复杂度.该算法在频率选择性慢衰落信道下,与LMMSE和低秩估计算法进行了相关性能及运算复杂度对比分析,结果表明,该算法能以较微弱的性能代价换取系统复杂度的明显降低.

关键词: 信道估计, 相关带宽, LMMSE算法, 低秩估计算法

Abstract:

This paper presents a novel low complexity LMMSE channel estimation algrithom for the OFDM system to reduce computational complexity caused by matrix inverse operation in the MMSE estimator. Correlation bandwidth is used as a criterion to divide the large auto-correlation matrix into a number of sub-matrixs, including the non-overlap and overlap method. The diaglog sub-matrix blocks representing low-frequency important channel information in the channel autocorrelation matrix are preserved while the other sub-matrix blocks are ignored, thus reducing autocorrelation matrix inversion computational complexity. BER and MSE performance of the new algrithom is evaluated in the frequency selective flat fading channel with comparison to the LMMSE, SVD algorithm. Simulation results and complexity anaysis show that low complexity is obtained at the slight cost of performance degradation.

Key words: channel estimation, coherence bandwidth, LMMSE, SVD

中图分类号: 

  • TN929.5