摘要
为提高传播算子算法在低信噪比下的波达方向(direction of arrival,DOA)估计性能,降低计算复杂度,提出了一种基于互相关矩阵的二维传播算子DOA估计实值算法(UC-PM).该算法通过构造新的互相关矩阵代替阵列接收数据矩阵,抑制了噪声分量的影响,并且保持了传播算子算法计算量小的优点,利用线性运算代替特征分解求得旋转不变关系矩阵.同时,为进一步降低算法计算量,利用酉变换思想构建新的实数域旋转不变关系,将特征分解和最小二乘问题实数化.仿真结果和计算复杂度分析表明,新算法在低信噪比下的估计性能优于传统二维传播算子算法,接近于二维ESPRIT算法,且其计算复杂度远小于二维ESPRIT算法,实时性好,具有良好的实用价值.
In order to improve the accuracy of direction of arrival（DOA）estimation and decrease calculation capacity with low SNR,a new real-valued propagator method（PM）for 2-D DOA estimation algorithm using cross-correlation matrix（UC-PM）was proposed.Instead of array received data,cross-correlation matrix was constructed to suppress the effect of noise,and eigendecomposition was replaced by a linear operator with low calculation capacity.Meanwhile,for reducing the complexity further,a new real-valued rotational invariance matrix was constructed to change eigen-decomposition and total least problems into real ones by unitary transformation.The simulation results showed that,being similar performance with 2-D ESPRIT algorithm,the performance of UC-PM is better than conventional PM algorithm with low SNR,and it has much lower calculation capacity than 2-D ESPRIT algorithm,which made the proposed algorithm of high practical value.
出处
《北京理工大学学报》
EI
CAS
CSCD
北大核心
2017年第1期71-76,共6页
Transactions of Beijing Institute of Technology
关键词
波达方向
互相关
传播算子
酉变换
direction-of-arrival
cross-correlation
propagator method
unitary transformation
