基于改进子空间追踪算法的稀疏信道估计

由于许多通信系统的信道具有稀疏多径的特性，因此可以将信道估计问题归结为稀疏信号的恢复问题，继而应用压缩感知理论(CS)的算法求解。针对CS中现存的信号重构方法——子空间追踪法(SP)需要对稀疏度有先验知识的缺点，提出一种改进的子空间追踪法(MSP)。该方法的反馈和精选过程与SP算法一致，不同之处是MSP算法每次迭代时向备选组合中反馈添加的向量个数是随着迭代次数而逐一增加的，而SP算法中备选组合被添加的向量个数与稀疏度相同。仿真结果表明，基于MSP方法所得到的稀疏多径信道估计结果优于基于传统SP的方法，且无需已知信道的多径个数。

Sparse channel estimation based on modified subspace pursuit algorithm

Due to the sparse structure of channels in a number of communication systems, the sparse channel estimation problem can be formulated as the reconstruction problem of sparse signals, and then being solved by certain algorithm in Compressive Sensing (CS) theory. To avoid needing prior knowledge for sparseness, a Modified Subspace Pursuit (MSP) was proposed. The feedback and refining processes of MSP are the same as those of the existing Subspace Pursuit (SP), the difference between them is that, in MSP, the number of vectors added to the candidate set is increased one by one, not equal to the number of sparseness in SP in every iteration. The simulation results show that, compared with the existing subspace pursuit method, the main innovative feature of the proposed algorithm is that it does not need to assume the sparseness of channel but offers superior estimation resolution.