MIMO-OFDM系统时域参数化信道估计算法研究

Parametric Channel Estimation Algorithm in Time Domain for MIMI-OFDM Systems

针对MIMO-OFDM系统信道估计问题,提出了一种在时域内进行最小二乘信道估计的方案.它的基本思想是利用发送端与接收端导频序列所形成的时域信号来计算信道的时域响应信息.为了获得LS信道估计的最小均方误差,本文给出了一种最优导频序列设计方案,它要求每根发射天线中的导频序列为等间隔排列,不同发射天线之间的导频序列位置相互交错.本文还提出了一种基于LS准则的参数化信道估计方法,对比传统的LS信道估计算法,本文算法能够大幅度地提高信道估计的精度;对比传统的LSPCE算法,本文算法可以有效地降低其计算复杂度.仿真实验和性能分析验证了所提方法的有效性.

For MIMO-OFDM channel estimation, a scheme of LS （least square） channel estimation in time domain is proposed. The basic idea of the scheme is that pilot signals of the transmitter and the receiver in time domain are used to calculate the channel response in time domain. To obtain the minimum mean square error （MMSE） in LS channel estimation, a design scheme for optimal pilot sequences is also proposed, and it is needed that not only the placement of pilot tones is of equal spacing in the frequency domain, but also the pilot positions are orthogonal to each other in different antennas. An LS-based parametric channel estimator （LSPCE） is also presented. The proposed algorithm gives much better performance than conventional LS algorithm and has lower computational complexity than conventional LSPCE. Simulation results and performance analysis validate the effectiveness of the proposed approach.