中文摘要：

该文提出一种非迭代的稀布线阵方向图综合方法。该方法首先对方向图采样数据进行centro-Hermit化处理,然后通过酉变换构造等价实矩阵束,得到非均匀单元位置与新矩阵束广义特征值的关系。在此基础上,对实矩阵奇异值分解,并舍弃非主要奇异值以获得低阶左奇异向量矩阵,进而求得稀布阵列的阵元位置和相应激励。相比于其他方法,该方法能够直接得到阵元位置的实数解,奇异值分解和特征值分解均在实数域进行,提高逼近程度的同时有效降低了计算量,仿真验证了该方法利用少量阵元即可高效实现线阵的方向图综合。

英文摘要：

A novel non-iterative method, named unitary matrix pencil method, is presented in this paper for the pattern synthesis of sparse linear arrays. Through unitary transformation of the centro-Hermit matrix constructed using sample data of the desired pattern, an equivalent real-valued matrix pencil can be achieved so as to determine the relation between non-uniform element positions and new generalized eigenvalues. Then, the lower order left singular vector matrix can be obtained by discarding the non-principal singular values generated by Singular Value Decomposition （SVD） of the real-valued matrix. The element positions and excitations are thereby estimated efficiently. Compared with other algorithms, this method can be utilized to directly obtain the real-valued solution of sparse array locations. Furthermore, Singular Value Decomposition （SVD） and Eigen Value Decomposition （EVD） are computed in the real-valued field with a lower computational cost. Simulation results validate the high-efficiency of the proposed synthesis method for the design of arbitrary linear array pattern with a fewer number of antenna elements.