Kalman Filter Channel Estimation in 2x2 and 4 x 4 STBC MIMO-OFDM Systems
Kalman Filter Channel Estimation in 2x2 and 4 x 4 STBC MIMO-OFDM Systems
复制标题
2 × 2 和 4 × 4 STBC MIMO-OFDM 系统中的卡尔曼滤波器信道估计
DOI：
10.1109/access.2020.3027377
复制
发表时间：
2020-01-01
期刊：
影响因子：
3.9
通讯作者：
Wang, Chengyou
中科院分区：
文献类型：
--
作者：
Tang, Ruiguang;Zhou, Xiao;Wang, Chengyou
信道估计对于动态环境中的空时分组编码（STBC）多输入多输出正交频分复用（MIMO-OFDM）系统来说是一个具有挑战性的问题。为了解决这个问题并提高系统性能，本文提出了一种基于卡尔曼滤波器（KF）的信道估计方法，应用于2 x 2和4 x4 STBC MIMO-OFDM系统。所提出的基于KF动态跟踪特性的方法很好地用于动态信道估计。首先，采用新的正交空时码字，设计正交导频序列来抑制发射天线间的干扰。然后，研究了KF的预测和更新特性，建立了STBC MIMO-OFDM系统的状态空间模型。随后，根据KF估计方程迭代估计信道状态信息(CSI)。最后，为了进一步提高KF信道估计的精度，利用阈值来抑制KF法估计的信道脉冲响应(CIR)中的噪声。仿真结果验证了所提出的正交导频和STBC的KF信道估计方法与其他传统信道估计方法相比提供了更好的误码率（BER）和归一化均方误差（NMSE）性能，并且可以有效地适应不同低阶和高阶调制的动态多径传播条件。
Channel estimation is a challenging problem for space time block coding (STBC) multipleinput and multiple-output orthogonal frequency division multiplexing (MIMO-OFDM) systems in dynamic environments. To handle this problem and improve the system performance, this paper proposes a Kalman filter (KF) based channel estimation method applied to 2 x 2 and 4 x4 STBC MIMO-OFDM systems. The proposed method based on the dynamic tracking property of KF is well adopted for dynamic channel estimation. First, a neworthogonal space-time codeword is adopted, and the orthogonal pilot sequences are designed to suppress the interference among transmit antennas. Then, the prediction and update characteristics of KF are researched, and the state space model is established for STBC MIMO-OFDM system. Subsequently, the channel state information (CSI) is estimated iteratively according to the KF estimation equation. At last, to further improve the accuracy of KF channel estimation, the threshold is utilized to suppress the noise in the channel impulse response (CIR) estimated by KF method. Simulation results verify that the proposed KF channel estimation method with orthogonal pilots and STBC provides better bit error rate (BER) and normalized mean square error (NMSE) performance compared with other conventional channel estimation methods, and it can be effectively adapted to dynamic multipath propagation conditions with different low and high order modulations.