EC Analysis, 7012 Clovercliff Drive, Rancho Palos Verdes, California 90275, USA.
J Acoust Soc Am. 2011 Feb;129(2):765-73. doi: 10.1121/1.3531811.
Conventional synthetic aperture processing uses motion of the sonar to increase aperture size and bearing resolution. Two recent papers discussed a different application in which synthetic elements are used to fill in an otherwise sparse passive array. This paper points out that ambiguities persist, even with synthetic elements, in the ideal case of a straight, uniform, sparse line array with constant velocity in the presence of plane wave signals. It is also shown that irregular motion such as acceleration introduces additional information which can be exploited to suppress the ambiguities. The degree of suppression in such an approach is independent of signal direction. If source stability supports extended coherent integration, then the acceleration and integration time required are both modest to achieve interesting levels of suppression. For a less stable source, a modified conventional beamformer is introduced which leverages acceleration over multiple snapshots to suppress the ambiguities. A post-beamformed processing stage involving a nonlinear deconvolution technique such as the CLEAN algorithm can further improve the result. A semi-coherent adaptation of CLEAN is shown to remove the residual ambiguities effectively in the presence of a moderate level of uncorrelated noise.
传统的合成孔径处理利用声纳的运动来增加孔径大小和方位分辨率。最近有两篇论文讨论了一种不同的应用，即在存在平面波信号的情况下，使用合成单元来填充稀疏的被动阵。本文指出，即使使用合成单元，在具有恒定速度的直的、均匀的、稀疏的线阵的理想情况下，仍然存在模糊。还表明，加速度等不规则运动会引入额外的信息，可以利用这些信息来抑制模糊。这种方法的抑制程度与信号方向无关。如果源稳定性支持扩展相干积分，则所需的加速度和积分时间都适中，可实现有趣的抑制水平。对于不太稳定的源，引入了一种修改后的传统波束形成器，它利用多个快照上的加速度来抑制模糊。涉及非线性反卷积技术（如 CLEAN 算法）的波束形成后处理阶段可以进一步提高结果。显示在存在中等水平的不相关噪声的情况下，半相干 CLEAN 自适应有效地消除了残余的模糊。
