基于压缩感知的MIMO-OFDM系统的信道估计

Channel Estimation of MIMO-OFDM System based On Compressive Sensing

由于多输入多输出正交频分复用(MIMO-OFDM)系统信道具有稀疏性特点,MIMOOFDM系统的信道估计问题就转变为稀疏信号的重建。传统的多输入多输出正交频分复用(MIMO-OFDM)无线通信系统的信道估计算法,没有充分利用无线信道固有稀疏性,导致信道估计精度和频谱资源利用率不高等问题。因此,提出一种可扩展的正交匹配追踪算法(Extended Orthogonal Matching Pursuit,OMP_α),α为扩展因子α∈[0,1]。在原有的正交匹配追踪算法(Orthogonal Matching Pursuit,OMP)的基础上,通过适当增加迭代次数(m+(αm))来选取更加精确的匹配原子,从而达到重构原信号的目的。仿真结果表明,与现有的OMP算法、最小二乘法(LS)相比,OMP_α算法提高了信号的重构概率和精度,同时也提高了频谱资源的利用率和信道估计的性能。

The channel estimation problem of MIMO-OFDM（Multiple-lnput Multiple-Output Orthogonal Frequency Division Multiplexing） system can be transformed into sparse signal reconstruction due to the sparsity of MIMO-OFDM system channel.The channel estimation algorithm of traditional MIMO-OFDM wireless communication system does not make full use of the inherently sparsity of wireless channel, which leads to that channel estimation accuracy and the rate of spectrttm resource utilization is not high. In this paper, we mainly adopts an extended orthogonal matching pursuit algorithm（OMPα） and a is the expansion factor（α∈[0,1]）.By appropriately increasing the number of iterations （ m ＋ [ am ] ） to select a more accurate matching of atoms based on the original OMP （Orthogonal Matching Pursuit）algorithm, so as to achieve the α of reconstructing the original signal.Simulation shows that OMPα algorithm not only improves the probability and precision of signal reconsauction compared with existing OMPα algorithm and the conventional LS （Least Square） channel estimation method but also improves the utilization of spectrum resources and the performance of channel estimation.