基于酉变换和稀疏贝叶斯学习的离格DOA估计

发布时间：2018-04-16 15:22

本文选题：到达角估计 + 酉变换　；参考：《通信学报》2017年06期

【摘要】：针对传统稀疏贝叶斯学习算法(SBL)在解决低信噪比条件下信号到达角(DOA)估计有效性的问题,提出基于酉变换的实数域稀疏贝叶斯学习(RV-OGSBL)的快速离格DOA估计方法。该方法首先对均匀线阵的实际接收信号通过构造增广矩阵作为DOA估计的处理信号,然后利用酉变换将估计模型从复数域转化到实数域,进一步在实数域下将离格模型与稀疏贝叶斯学习算法相结合迭代处理实现DOA估计,获得较高的估计精度。仿真结果表明,RV-OGSBL方法不仅能保持传统SBL算法的性能,而且显著降低了计算复杂度。在低信噪比和低快拍数的情况下,算法运行时间降低约50%,表明该方法是一种快速的DOA估计算法。
[Abstract]:In order to solve the problem of the validity of the traditional sparse Bayesian learning algorithm (SBL) in solving the problem of signal arrival angle (OAA) estimation under low signal-to-noise ratio (SNR), a fast DOA estimation method based on unitary transform for sparse Bayesian learning (RV-OGSBL) in real number domain is proposed.In this method, the real received signal of uniform linear array is first processed by constructing an augmented matrix as the DOA estimation signal, and then the estimation model is transformed from complex number domain to real number domain by unitary transformation.Furthermore, in real number domain, the DOA estimation is realized by the combination of the discrete model and sparse Bayesian learning algorithm, and the estimation accuracy is higher.Simulation results show that the RV-OGSBL method can not only maintain the performance of the traditional SBL algorithm, but also significantly reduce the computational complexity.Under the condition of low signal-to-noise ratio (SNR) and low beat number, the running time of the algorithm is reduced by about 50%, which indicates that this method is a fast DOA estimation algorithm.
【作者单位】：上海大学通信与信息工程学院;
【基金】：上海“东方学者”专项基金资助项目(No.B60-D107-14-201)~~
【分类号】：TN911.7

【相似文献】