基于GOMP及其改进的OFDM系统稀疏信道估计

Generalized Orthogonal Matching Pursuit and Improved Algorithms for Compressive Sensing Based Sparse Channel Estimation in OFDM Systems

摘要: 研究在正交频分复用（OFDM）系统的稀疏信道估计问题.由于在许多通信系统中信道具有稀疏性，因此可以把信道估计问题转化为稀疏信号的恢复问题，应用压缩感知理论求解，把现有的恢复算法——广义正交匹配追踪算法（GOMP）运用到信道估计中，并对它加以改进.仿真结果表明，与广义正交匹配追踪算法（GOMP）相比，正交匹配追踪算法（OMP）运行时间少，计算复杂度低，但是估计的最小均方误差略差.为了进一步提高该算法的性能，提出了改进的广义正交匹配追踪算法，性能得到了较大的提高.

Abstract: The sparse channel estimation in orthogonal frequency division multiplexing (OFDM) systems was studied to solve the channel sparsity problem in many communication systems. The sparse channel estimation problem was formulated as the reconstruction problem of sparse signals. Based on compressive sensing theory, generalized orthogonal matching pursuit (GOMP) was applied to the channel estimation, and an improved GOMP algorithm was proposed. Simulation results demonstrate that, compared with OMP, though GOMP brings in some sort mean square error (MSE), but it shows lesser running time and lower computational complexity. And the effect of improved GOMP algorithm is better for the performance improvement of GOMP.