Sparse Sensor Array Design and Processing

阵列信号处理利用多个自适应控制的传感器，在各种传感和通信系统中起着基础性的作用。阵列信号处理是实现自适应波束形成、干扰抑制、多用户接入、波达方向(DOA)估计、信源定位和成像的关键技术。阵列信号处理技术广泛应用于无线通信、雷达、声纳、医学成像和射电天文学等领域。现代传感和通信应用对空间分辨率和估计精度的要求越来越高，这就需要更大的阵列孔径。传统的均匀线阵的传感器放置在奈奎斯特半波长间隔，要求传感器的数量与阵列孔径近似成正比。该项目开发了新的稀疏阵列设计策略和信号处理技术，以提供经济高效的解决方案，以实现大阵列孔径和高空间分辨率，并使用更少的传感器。特别是，它将开发新的稀疏阵列设计策略，包括基于结构化的阵列内插并减少相关滞后域中的冗余。这些解决方案将带来多重好处，如降低下一代无线通信的硬件复杂性和增强通信和传感性能，支持用于自动驾驶的高分辨率汽车雷达成像，以及促进优化的可重构智能表面应用。在教育方面，该项目将提供培训研究生和本科生的机会，支持当地高中生的暑期研究活动，并从代表性不足的群体中招收学生。拟议的努力通过以下三个方面加以整合。(A)在第一个推力中，发展了新的稀疏阵列设计策略和信号处理技术。稀疏阵列的设计和优化是为了最大限度地减少相关滞后中的冗余，并通过考虑阵列内插能力来获得高自由度。信号带宽被利用来提供频率分集，从而扩展了阵列的空间维度并分辨了更多的源。(B)在第二个推力中，考虑了相关和相干信号的DOA估计的挑战性问题。这项工作将检验稀疏阵列的内插能力，分析可实现的自由度，并为其实现设计有效的信号处理技术。(C)在第三个推力中，考虑到许多实际场景中信号到达的空间稀疏性，提出了压缩测量策略来降低大尺寸阵列信号的维度，从而显著降低了用于数字处理的采样信号的维度。从信息论、机器学习和优化的角度设计了新的压缩测量方法，以提高DOA估计的性能。该项目的成果将推动稀疏阵列的概念、理论和算法超越当前最先进的水平，并将拓宽稀疏阵列信号处理技术在实践中的适用性和健壮性。该奖项反映了NSF的法定使命，并通过使用基金会的智力优势和更广泛的影响审查标准进行评估，被认为值得支持。

Array signal processing exploiting multiple adaptively controlled sensors plays a fundamental role in various sensing and communication systems. Array signal processing is a key enabling technology for adaptive beamforming, interference suppression, multiuser access, direction-of-arrival (DOA) estimation, source localization, and image formation. Array signal processing techniques are broadly utilized in various applications including wireless communications, radar, sonar, medical imaging, and radio astronomy. The demand for increasingly higher spatial resolution and estimation accuracy in modern sensing and communication applications calls for larger array apertures. Conventional uniform linear arrays with sensors being placed at the Nyquist half-wavelength spacing require the number of sensors be approximately proportional to the array aperture. This project develops new sparse array design strategies and signal processing techniques to offer cost-effective solutions that enable large array apertures and high spatial resolutions with a smaller number of sensors. In particular, it will develop novel sparse array design strategies that incorporate structured-based array interpolations and reduce redundancies in the correlation-lag domain. Such solutions will bring multifold benefits, such as reducing the hardware complexity and enhancing communication and sensing performance in next-generation wireless communications, supporting high-resolution automotive radar imaging for autonomous driving, and facilitating optimized reconfigurable intelligent surface applications. On the educational front, this project will offer opportunities for training of graduate and undergraduate students, supporting summer research activities of local high school students, and recruiting students from underrepresented groups. The proposed efforts are integrated through the following three thrusts. (a) In the first thrust, new sparse array design strategies and signal processing techniques are developed. Sparse arrays are designed and optimized to minimize redundancies in the correlation lags and obtain a high number of degrees of freedom by accounting for the array interpolation capability. Signal bandwidth is exploited to offer frequency diversity that extends the array spatial dimension and resolves more sources. (b) In the second thrust, the challenging problem of estimating the DOAs of correlated and coherent signals is considered. The proposed work will examine the sparse array interpolation capability, analyze the achievable degrees of freedom, and devise effective signal processing techniques for their implementations. (c) In the third thrust, recognizing the spatial sparsity of signal arrivals in many practical scenarios, compressed measurement strategies are developed to reduce the dimension of large-size array signals so the signal dimension to be sampled for digital processing is significantly reduced. New compressed measurement methods are devised from information-theoretical, machine learning, and optimization perspectives to enhance DOA estimation performance. The outcomes from this project will advance the sparse array concept, theory, and algorithms beyond the current state-of-the-art, and will broaden the applicability and enhance the robustness of sparse array signal processing techniques in practice.This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.