OFDM系统中基于弱选择ROMP算法的稀疏信道估计

为了解决实际OFDM通信系统中信道稀疏度未知的不足，提出将弱选择正则化正交匹配追踪算法用于估计稀疏信道。算法在不知晓信道稀疏度的情况下，对不同迭代残差与测量矩阵中原子的相关系数进行判定后，根据原子的弱选择准则灵活地确定出表示信道冲激响应的原子候选集，进而利用正则化原则从候选集中挑选出表示信道冲激响应的最优原子组，逐步实现精确重建。仿真结果和理论分析表明：与正则化正交匹配追踪算法相比，相同条件下改进算法可以获得更低的均方误差和误比特率；另外，算法无需将信道稀疏度作为先验信息，实用性更强。

OFDM sparse channel estimation based on weak selected regularized orthogonal matching pursuit algorithm

To overcome the drawback that the sparsity of channel impulse response is usually unknown in existing Orthogonal Frequency Division Multiplexing(OFDM) systems,a weak selected regularized orthogonal matching pursuit algorithm is used for estimating sparse channel. Without the knowledge of the sparsity of channel,the algorithm judges the relevance coefficients between different iterative residue and atoms of measurement matrix,determines the number of atoms and the candidate atoms of the channel impulse response flexibly on the basis of a weak selection criterion,and then selects an optimal atom set for the channel impulse response from the candidate atoms according to the regularization principle,thus achieving accurate reconstruction progressively. Simulation results and theoretical analysis show that,compared with regularized orthogonal matching pursuit(OMP) algorithms,the proposed algorithm gets a much lower mean square error and bit error rate under the same condition;besides,as the algorithm does not need the channel sparsity as a priori information,it is much suitable for real application.