论文部分内容阅读
针对双基地多输入多输出(MIMO)雷达目标数估计问题,提出一种基于三阶张量分解的目标数估计算法。该算法首先将匹配滤波输出转化为三阶张量;然后,从张量的三个维度分别计算特征值,并通过迭代方法估计全局特征值;最后,利用全局特征值估计目标数。与传统基于矩阵分解算法相比,该算法充分利用了回波数据的多维结构特征,可有效改善特征值的估计精度,从而提高目标数的正确估计概率。仿真结果证明了所提算法的有效性和优越性。
Aiming at the problem of target number estimation in bistatic multiple input multiple output (MIMO) radar, a target number estimation algorithm based on third-order tensor decomposition is proposed. Firstly, the matched filter output is transformed into the third-order tensor. Then, the eigenvalues are calculated respectively from the three dimensions of the tensor, and the global eigenvalues are estimated by iterative method. Finally, the global eigenvalues are used to estimate the target numbers. Compared with the traditional algorithm based on matrix factorization, this algorithm makes full use of the multidimensional structure of echo data, which can effectively improve the estimation accuracy of eigenvalue and improve the probability of correct estimation of target number. The simulation results prove the effectiveness and superiority of the proposed algorithm.