基于矩阵填充的合成宽带高频雷达非网格目标分辨技术研究

发布时间：2018-05-19 18:54

本文选题：高频雷达 + 目标分辨　；参考：《电子与信息学报》2017年12期

【摘要】：高频雷达由于工作在拥挤的高频频段(3~30 MHz),连续频带资源十分有限,有限的带宽限制了对目标的分辨能力。该文引入一种合成宽带的发射信号,同时针对该信号,提出一种基于矩阵填充(Matrix Completion,MC)的1维和2维目标参数估计方法,分别称之为矩阵填充1维估计(MCE-1D)和矩阵填充2维估计(MCE-2D)方法,该方法将不完备采样集合变换成低秩矩阵,通过构造双重汉克尔(two-fold Hankel)矩阵将其转化为半定规划(SemiDefinite Programming,SDP)问题求解。新方法应用于高频雷达中,可以在非连续谱的背景下获得场景中目标位置的准确估计,很好地解决了非网格目标在传统网格类方法中的基失配问题,新方法对于非网格目标具有更高的分辨能力及抗噪性能。仿真处理结果验证了该方法的有效性。
[Abstract]:Because the HF radar is working in the crowded high frequency band, the continuous frequency band resource is very limited, and the limited bandwidth limits the resolution ability of the target. In this paper, a synthetic wideband transmitting signal is introduced. For this signal, a method for estimating 1-D and 2-D target parameters based on matrix filling matrix is proposed. The method is called matrix filled 1-dimensional estimator MCE-1D) and matrix filled 2-dimensional estimator / MCE-2D) method. The incomplete sampling set is transformed into a low-rank matrix by constructing a two-fold Hankel-based matrix, which is transformed into a semidefinite programming semifinite programming SDP problem by constructing a two-fold Hankel-based matrix. The new method is applied to high frequency radar, which can obtain the accurate estimation of the target position in the scene under the background of discontinuous spectrum, which solves the base mismatch problem of the non-grid target in the traditional mesh class method. The new method has higher resolution and anti-noise performance for non-grid targets. The simulation results show that the method is effective.
【作者单位】：哈尔滨工业大学电子与信息工程学院;哈尔滨工业大学信息感知技术协同创新中心;
【基金】：国家自然科学基金(61171182,61032011,61171180,61571159) 中央高校基本科研业务费专项资金(HIT.MKSTISP.201613,HIT.MKSTISP.201626)~~
【分类号】：TN957.51

【相似文献】