稀疏OFDM信道估计的同伦算法

Homotopy Algorithm for Sparse OFDM Channel Estimation

针对现有稀疏信道估计面临的复杂度过高、性能不佳、未充分利用稀疏性等缺陷,提出了联合压缩感知和同伦法的稀疏信道估计.主要通过跟踪正则因子,不断更新解方向和迭代步进,从而获得解的更新方程,具有复杂度低,性能好等优势.通过仿真及分析得到:在低信噪比条件下,与传统最小二乘法等算法比较,所提基于压缩感知同伦法的稀疏信道估计算法有20dB性能增益,获得了较好估计性能.

Current sparse channel estimation is mainly confronted with the deficiencies, such as high complexity, poor performance, not adequately exploiting the sparse property, and so on. In order to solve these problems, a joint compressive sensing and the homotopy method is proposed to estimate the sparse channel. By tracking the change of regular factors, it constantly updates the direction of the solution and the iterative step. So it can obtain the solution of the update equation. It also has the advantage of low complexity, good performance and so on. The simulation channel estimation is better than the traditional least squares（LS） algorithm. At low signal-to-noise ratio（SNR）, the proposed algorithm can obtain about 20 dB performance gain than that of the traditional algorithms. And the computational complexity is also reduced to 1/2 of that of the traditional algorithms. Therefore, it obtains a rather better estimation performance.