中文摘要：

提出了一种基于重要性采样方法的最大似然多径信道估计方法。该方法将时延与幅度估计解耦,利用蒙特卡罗算法（MC）对未知参数的分布函数抽样,计算样本均值直接得到多径时延和幅度估计结果,避免了耗时较长的多维网格搜索和对初值较为敏感的迭代算法。仿真结果表明,对于多径路数分别为5、15的情况,均能获得较好的信道估计。相比于现有的其他一些算法,该算法不仅能够同时获得时间延迟估计,同时也能对幅度精确估值。

英文摘要：

Based on the importance sampling （IS）, this paper proposes a new algorithm to estimate the time delay and amplitude in multipath channels. This algorithm decouples the time delay and amplitude estima- tion and utilizes Monte Carlo method for sampling unknown parameters. By so doing the parameters are approximated by the average samples, and the estimator proposed in this paper avoids the complex mul- tidimensional grid search and iterative methods which depend seriously on the initial guess. The results show that the algorithm performs well under condition of respective multi-paths 5, 15. Compared with some other known algorithms under the same condition, the algorithm can achieve not only the time delay estimation, but also a precision amplitude estimation simultaneously.