摘要
基于对自适应滤波算法的研究可以得出,GSC(Generalized Sidelobe Canceller)框架是所有降秩滤波算法的统一基础模型,通过对GSC一般结构的降秩模型的研究。文中提出了3种能够改善一般结构结果的优化模型,PC(Prin-ciple Component)主分量算法、CS(Cross Spectrum)交叉谱算法、MWF(Mutistage Wiener Filter)多级维纳滤波器。这三种方法都是基于特征子空间截断的方法。通过对降秩矩阵T的特征值分解和重新构造,大大降低了自由度和工程运算量。该两种方法在实际应用中具有更优的实时性。仿真结果证明了提出的基于广义旁瓣相消的降秩自适应波束形成算法具有良好的降低自由度和波束形成性能,验证了算法的有效性。
The research based on algorithms for reduced rank adaptive filtering indicates that the GSC Framework is the basic and uniform model around all kinds of algorithms for reduced-rank adaptive filtering with the research about general GSC for reduced-rank transformation. This paper demonstrates two measures about optimized model in order to improve the results of general framework, Principle Component,Cross Spectrum,and Multistage Wiener Filter, both of which are based the truncation of the Kry-lov subspace, which can dramatically decrease the degree of freedom and need little complex additions through eigendecomposition and reforming of matrix T. These two methods have more excellent real-time responsibility. The result of simulation proves that the GSC framework about reduced-rank algorithm has good beam-forming performance in dealing with decreasing degree of freedom, and the validity of these methods are indicated.
出处
《电子设计工程》
2013年第5期61-64,共4页
Electronic Design Engineering
基金
航空科学基金(2010ZC53028)
西北工业大学种子基金(Z2012091)
