An Exact Near-Field Model Based Localization for Bistatic MIMO Radar With COLD Arrays
An Exact Near-Field Model Based Localization for Bistatic MIMO Radar With COLD Arrays
复制标题
DOI：
10.1109/tvt.2023.3294625
复制
发表时间：
2023-12
影响因子：
6.8
通讯作者：
Hua Chen;Weilong Wang;Wei Liu;Ye Tian;Gang Wang
中科院分区：
文献类型：
--
作者：
Hua Chen;Weilong Wang;Wei Liu;Ye Tian;Gang Wang
现有的近场源定位算法大多基于菲涅耳近似模型，并假设目标在各传感器处的空间振幅相等。与这些算法不同，NF源参数估计算法，提出了基于精确的空间传播几何模型，双基地多输入多输出（MIMO）雷达部署与线性同心正交环和偶极子（COLD）阵列在发射机和接收机。该方法首先将接收端匹配滤波器的输出信号压缩成三阶并行因子（PARAFAC）数据模型，然后对其进行三线性分解，得到三个因子矩阵。然后，利用旋转不变性和Khatri-Rao乘积，从空间幅度比中估计多个感兴趣的参数，包括出发方向（DOD）、到达方向（DOA）、发射机到目标的距离（RFTT）、目标到接收机的距离（RFTR）、二维（2-D）发射极化角（TPA）和二维接收极化角（RPA）。最后，发射和接收阵列的相位不确定性可以从附加的相位项中提取。该算法避免了谱峰搜索，并且可以自动无歧义地匹配封闭形式的估计参数。此外，它适用于任意阵元间距和相位不确定性的非均匀线阵（NLA）。仿真结果证明了该方法的优越性。
Most existing near-field (NF) source localization algorithms are developed based on the Fresnel approximation model, and assume that the spatial amplitudes of the target at the sensors are equal. Unlike these algorithms, an NF source parameter estimation algorithm is proposed, based on the exact spatial propagation geometry model, for bistatic multiple-input multiple-output (MIMO) radar deployed with a linear concentered orthogonal loop and dipole (COLD) array at both the transmitter and receiver. The proposed method first compresses the output signal of the matched filter at the receiver into a third-order parallel factor (PARAFAC) data model, on which a trilinear decomposition is performed, and subsequently three factor matrices can be obtained. Then, multiple parameters of interest, including direction-of-departure (DOD), direction-of-arrival (DOA), range from transmitter to target (RFTT), range from target to receiver (RFTR), two-dimensional (2-D) transmit polarization angle (TPA) and 2-D receive polarization angle (RPA), are estimated from the spatial amplitude ratio exploiting the rotation invariant property and the Khatri-Rao product. Finally, the phase uncertainties of transmit and receive arrays can be extracted from additional phase items. The proposed algorithm avoids spectrum peak search, and the estimated parameters in closed forms can be automatically matched unambiguously. In addition, it is suitable for non-uniform linear arrays (NLA) with arbitrary array element spacing and phase uncertainty. Advantages of the proposed method are demonstrated by simulation results.