一种改进的特征子空间平滑技术研究

Research on an ameliorated spatial smoothing technology based on eigenspace

近年来空间平滑技术在减小有限次采样引起的误差、相关信号解等相关方面取得了巨大的进步,此算法与采样协方差矩阵求逆(DMI)法相比,可使信干噪比有一定提高,但改善程度有限.针对空间平滑技术这一特点提出了一种新的波束形成算法,并进行了理论分析.该算法通过对特征值门限的判定,使得在期望信号较小时,采用改进的权矢量,从而达到了在较小快拍数情况下仍能有效地提高信干噪比,并通过性能仿真验证了该算法的有效性.

In recent years, the technology of spatial smoothing has made great progress in the aspect of diminishing errors caused by the limited time of samples, and decorrelation of relevance signals. This algorithm can do better than DMI in improving SINR, but the degree of improvement is limited. This paper advances a new beam-forming algorithm aiming at the characteristic of spatial smoothing technology, and carries out theoretical analysis. By deciding on the characteristic value threshold, this algorithm adopts the ameliorated power vector when the expectation signal is less, thereby still improving SINR under few snapshots situation, and performance simulation verifies the validity of this algorithm.