正交频分复用-多天线系统的盲信道估计算法

为了解决多天线系统的识别模糊问题,提出一种基于系统输出信号二阶统计的正交频分复用多天线(OFDM-MIMO)系统盲信道估计算法。该算法从频域角度出发,从而将有限冲击响应信道的识别简化为一系列的无记忆系统识别问题。利用信道频率相关特性降低了算法复杂度,同时解决了算法对样本数的依赖问题。为了去除盲识别的模糊度,设计了带少许导频插入的OFDM-MIMO系统结构,研究了算法的估计精度及其对信道容量的影响。仿真结果表明:2发2收,1024子载波的OFDM-MIMO系统在信噪比为20dB时盲估计的归一化均方误差约为10-2。另外,子载波信道的容量结果表明,在信道变化较快时该文提出的盲估计算法相对于训练方法有明显的容量优势。

Blind channel estimation algorithm for OFDM-MIMO systems

In order to solve the identification confusion in multi-input-multi-output (MIMO) systems, an efficient blind channel estimation algorithm based on second-order statistics was developed for orthogonal frequency division multiplexing multi-input-multi-output (OFDM-MIMO) systems. The algorithm analyzes the system in the frequency domain, then simplifies the finite-impulse-response channel estimate into several memoryless systems. The algorithm uses the channel correlation in the frequency domain to greatly reduce the computation complexity and the number of required samples. The OFDM-MIMO structure with pilot insertion eliminates the ambiguity of blind identification. Analysis of the estimation accuracy and the subcarrier channel capacity shows that the normalized mean square error (NMSE) of the algorithm is about 10~(-2) for a 2-input-2-output OFDM-MIMO system with 1 024 subcarriers and that the blind algorithm helps the system to achieve more capacity than that of training based systems in fast-varying channels.