【摘要】 在船舶自动识别系统中,波束形成算法在目标信号受到来向相近的部分重叠干扰时存在性能严重下降的问题。针对上述问题,提出一种超分辨率下的部分重叠干扰抑制算法。算法通过自回归模型预测的方法求解自回归系数,进而进行虚拟阵元扩展预处理,并采用基于广义奇异值分解的波束形成算法完成目标信号的估计以及对部分重叠干扰的抑制,克服了上述场景下波束形成器在信号来向相近时性能较差的问题。仿真结果表明,所提算法通过结合阵列扩展与波束形成的特点,能够在不增加实际阵元的情况下较原阵列具备更高的分辨率。更多还原

【Abstract】 In order to solve the problem in Automatic Identification System that the performance of beamforming algorithm is seriously degraded when the target signal is subjected to partial overlapping signal interference from the near-incident direction, a partially overlapped interference suppression algorithm in super-resolution is presented. Firstly, the auto-regressive coefficient was calculated with auto-regressive model and performs the virtual array extrapolation. Then, the beamforming algorithm based on generalized singular value decomposition was utilized to estimate the target signal and suppress the partially overlapped interference. The weak capacity against signals from the closely incident angle in beamformer was strengthened. Simulation results show that the proposed algorithm based on the combination of array extrapolation and beamforming can effectively achieve the higher resolution than the original array without increasing the elements of actual array.更多还原