基于块稀疏恢复的空时自适应信号处理研究

发布时间：2018-05-29 21:29

本文选题：空时自适应处理技术 + 压缩感知　；参考：《南京理工大学》2017年硕士论文

【摘要】：空时自适应处理(STAP)技术利用空域和时域的信息有效抑制了杂波,与稀疏恢复算法相结合可以减少所需的样本数据,但是,在样本数量严重不足时,该方法恢复出的杂波只能得到大概位置,与真实谱相差甚远。由于STAP信号中的非零值不仅具有稀疏特性,而且显著值还具有聚类的特性,即具有块稀疏特性,因此可以考虑利用STAP信号特有的块稀疏特性恢复杂波谱,从而提高抑制杂波的性能。本文将块稀疏重构理论应用于STAP技术,在样本量严重不足时,较大地提高了重构精度与杂波抑制性能,并且计算时间在可接受范围内。本文所做的工作具体如下:1.介绍STAP的基本原理,分析杂波谱特性与性能指标。同时介绍了信号稀疏重构的原理与方法,选择正交匹配追踪(OMP)算法与光滑l0(SLO)算法这两种经典方法梳理算法过程,并且将这两种算法应用于STAP技术中,分析其性能上的不足。2.介绍块稀疏重构的原理,并在分析了 STAP杂波空时谱具有块稀疏特性的基础上,创新性地将块稀疏重构算法与STAP技术相结合,提出了基于块稀疏恢复算法重构STAP杂波谱的算法流程,并对算法步骤进行了详细研究。3.将OMP算法推广为块稀疏情况下的块正交匹配追踪(BOMP)算法并通过重构STAP杂波谱仿真分析得出,在BOMP应用于STAP技术时,由于STAP杂波子块边界未知,块稀疏分块的不准确会使贪婪算法本身易陷入局部最优解的缺陷放大。因此本文提出了一种BOMP修正算法,对要选择的最优原子块进行修正,设能量阈值对该子块边界进行判断,避免上述的重构误差,提高了重构精度与抑制杂波性能。同时,用Matlab仿真数据和MountainTop实测数据分别重构杂波谱进行比较,验证了提出算法的有效性。4.将SL0算法推广为块稀疏情况下的块光滑l0(BSL0)算法并用Matlab仿真数据和MountainTop实测数据分别重构杂波谱进行比较分析并验证:用块稀疏恢复算法处理STAP信号性能要优于普通稀疏恢复算法应用于STAP的性能。同时,在相同条件下横向比较两类算法,证明上述结论。
[Abstract]:The space-time adaptive processing (STAP) technique can effectively suppress clutter by using spatial and temporal information. Combining with sparse recovery algorithm, the required sample data can be reduced. However, when the number of samples is seriously insufficient, The clutter recovered by this method can only get the approximate position, which is far from the true spectrum. Because the non-zero values in STAP signal have not only sparse characteristics, but also significant values have clustering characteristics, that is, block sparsity, so we can use the block sparsity characteristic of STAP signal to recover clutter spectrum. In order to improve the performance of clutter suppression. In this paper, the block sparse reconstruction theory is applied to STAP technology. When the sample size is seriously insufficient, the reconstruction accuracy and clutter suppression performance are greatly improved, and the computation time is within the acceptable range. The work done in this paper is as follows: 1. The basic principle of STAP is introduced, and the characteristics and performance of clutter spectrum are analyzed. At the same time, the principle and method of signal sparse reconstruction are introduced. Two classical algorithms, orthogonal matching tracking algorithm (OMP) and smooth l0sloo algorithm, are selected to sort out the process, and the two algorithms are applied to STAP technology to analyze their performance deficiency. 2. This paper introduces the principle of block sparse reconstruction, and on the basis of analyzing the block sparse characteristic of STAP clutter space-time spectrum, innovatively combines block sparse reconstruction algorithm with STAP technology. The algorithm flow of reconstruction of STAP clutter spectrum based on block sparse recovery algorithm is proposed, and the algorithm steps are studied in detail. The OMP algorithm is extended to block orthogonal matching tracking (Bomp) algorithm in the case of block sparsity, and the simulation analysis of reconstructed STAP clutter spectrum shows that when BOMP is applied to STAP technology, the STAP clutter sub-block boundary is unknown. The inaccuracy of block sparse partitioning makes the greedy algorithm itself prone to the defect amplification of the local optimal solution. In this paper, a BOMP correction algorithm is proposed, which modifies the optimal atomic block to be selected, determines the boundary of the sub-block by energy threshold, avoids the reconstruction error mentioned above, and improves the reconstruction accuracy and clutter suppression performance. At the same time, the Matlab simulation data and the MountainTop measured data are compared to reconstruct the clutter spectrum respectively, which verifies the effectiveness of the proposed algorithm. 4. In this paper, the SL0 algorithm is extended to block smooth l0 BSL0 algorithm in the case of block sparsity. By comparing and analyzing the reconstructed clutter spectrum with Matlab simulation data and MountainTop measured data, the performance of block sparse restoration algorithm in STAP signal processing is better than that in common STAP signal processing. Sparse recovery algorithm is applied to the performance of STAP. At the same time, the above conclusion is proved by comparing two kinds of algorithms horizontally under the same conditions.
【学位授予单位】：南京理工大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TN911.7

【相似文献】