改进SSMUSIC超分辨多径时延估计算法

发布时间：2018-05-15 14:01

本文选题：超分辨时延估计 + 多重信号分类　；参考：《信号处理》2014年08期

【摘要】：传统的MUSIC超分辨时延估计技术是直接基于测量数据,其性能往往只对宽带且频谱近似平坦的信号较优,而对窄带信号估计性能较差。针对上述问题,本文通过利用谐波频率估计模型和DOA(Direction of Arrival)估计模型之间的等价性,将时延估计问题转化为谐波频率估计问题,提出了一种改进SSMUSIC(Signal Subspace Scaled Multiple Signal Classification)超分辨多径时延估计算法。改进后的算法采用平滑的思想和SSMUSIC算法的思想构造协方差矩阵和MUSIC谱,实现了对多径时延的超分辨估计。仿真表明,该算法能够实现对窄带信号多径时延超分辨估计且具有DP(Direct-Path)不模糊和谱峰陡峭的特点,估计性能优于传统的超分辨算法。
[Abstract]:The traditional MUSIC super-resolution time delay estimation technique is based on measured data directly. Its performance is usually better for wideband signals with approximately flat spectrum, but poor for narrowband signals. In view of the above problems, by using the equivalence between the harmonic frequency estimation model and the DOA(Direction of Arrival estimation model, the problem of time delay estimation is transformed into the harmonic frequency estimation problem. An improved SSMUSIC(Signal Subspace Scaled Multiple Signal Classification) super-resolution multipath time delay estimation algorithm is proposed. The improved algorithm uses the idea of smoothing and the idea of SSMUSIC algorithm to construct covariance matrix and MUSIC spectrum to realize the super-resolution estimation of multipath delay. Simulation results show that the proposed algorithm can realize the multipath delay super-resolution estimation of narrowband signals, and has the characteristics of DPG-Direct-Path-nonblur and steep spectral peak. The performance of the proposed algorithm is superior to that of the traditional super-resolution algorithm.
【作者单位】：解放军信息工程大学信息与系统工程学院;
【基金】：国家自然科学基金(61201381) 信息工程学院未来发展基金资助项目(YP12JJ202057)
【分类号】：TN911.23

【相似文献】