基于压缩感知的智能天线 DOA 估计算法

波达方向（direction of arrival，DOA）估计是利用具有一定空间结构的天线阵元对空间信源的来波方向进行估计。为解决MUSIC算法和ESPRIT算法在智能天线DOA估计中空间谱分辨率和抗噪声性能方面的不足，依据压缩感知理论的贪婪算法思想，将匹配追踪（matching pursuit，MP）算法和子空间追踪（subspace pursuit，SP）算法应用在DOA估计中，即提出2种新的算法MP-DOA和SP-DOA。这2种算法主要包括原子库建立、信号投影、最佳匹配原子筛选及信号特征提取等过程。4种算法的仿真实验与对比分析结果表明，MP-DOA算法在低信噪比情况下较其他3种算法性能更优越，SP-DOA算法在角度分辨率和对信源数的包容方面的性能较其他3种算法优越，也验证了应用压缩感知理论的优越性。

DOA estimation algorithms in smart antennas based on compressed sensing

Direction of arrival(DOA) estimation is realized by using antenna array with a certain spatial structure to estimate the arrival direction of space signal source. In order to solve the shortcomings in spatial spectrum resolution and antinoise property of MUSIC algorithm and ESPRIT algorithm used in DOA estimation in smart antennas, based on the theory of compressed sensing basic idea of the greedy algorithm, the matching pursuit (MP)algorithm and subspace pursuit (SP) algorithm are applied in DOA estimation. Two new algorithms including MP-DOA and SP-DOA are presented, which mainly include the processes of the establishment of dictionary, the projection of signals, the screening of the best matching atom and the feature extraction of signals. The simulation experiment and analysis indicate that the MP-DOA algorithm is superior to other three algorithms in low SNR, and SP-DOA algorithm is superior to other three algorithms in angular resolution and number of signal sources inclusive. The superiority of compressed sensing theory is also verified.