萬(wàn)連城 黑蕾 王迎斌
關(guān)鍵詞: 二維DOA估計(jì); 壓縮感知; 貝葉斯; 多任務(wù)貝葉斯壓縮感知; 分辨率; 算法復(fù)雜度
中圖分類號(hào): TN951?34 ? ? ? ? ? ? ? ? ? ? ? ? 文獻(xiàn)標(biāo)識(shí)碼: A ? ? ? ? ? ? ? ? ? ? ? 文章編號(hào): 1004?373X(2019)06?0010?04
Abstract: The constant development of the compressed sensing theory provides a new idea for the problem of 2?D direction of arrival (DOA) estimation. The traditional 2?D DOA estimation method is only the extension of the 1?D DOA estimation, and the modeling method of the 2?D DOA estimation is the same as that of the 1?D DOA estimation, which leads to problems of high computation complexity and low resolution in solving. The multitask Bayesian compressive sensing (MT?BCS) theory is applied to the 2?D DOA estimation problem by remodeling of the 2?D DOA model, so as to propose a separable 2?D DOA estimation algorithm based on MT?BCS. A comparative experiment was carried out. The results demonstrate that the proposed algorithm has the advantages of high resolution and low complexity.
Keywords: 2?D DOA estimation; compressed sensing; Bayesian; MT?BCS; resolution; algorithm complexity
基于稀疏表示[1?3]的二維DOA(Direction of Arrival)估計(jì)算法大多是基于一維DOA估計(jì)的擴(kuò)展,算法建模時(shí)也是將二維矩陣展開(kāi)為向量,仿照一維DOA估計(jì)的建模方法進(jìn)行建模。這類算法主要有:基于[lp]范數(shù)的POCUSS算法[2?4],經(jīng)典的高分辨[lp?SVD]算法[5],MP[6],OMP[7?8]等貪婪算法和基于貝葉斯壓縮感知的DOA估計(jì)算法[9]。
然而,這類仿照一維DOA的二維DOA建模方法導(dǎo)致稀疏基矩陣的維度過(guò)大,求解時(shí)算法的時(shí)間復(fù)雜度過(guò)高,難以滿足實(shí)時(shí)性的要求。為了降低算法的時(shí)間復(fù)雜度,本文提出了可分離的二維DOA建模新方法,并使用MT?BCS(Multitask Bayesian Compressive Sensing)算法[10]進(jìn)行求解,成功解決了二維DOA估計(jì)算法時(shí)間復(fù)雜度高、分辨率低的缺點(diǎn)。
由表1可知,由于本文所提出的方法將矩陣[A∈CML×PQ] 分離為俯仰維導(dǎo)向矢量基矩陣[Ψ∈CM×L]和方位維[Ψ]導(dǎo)向矢量基矩陣[Θ∈CP×Q],從而有效地減少了算法的時(shí)間復(fù)雜度,使算法更適合工業(yè)應(yīng)用。
對(duì)于傳統(tǒng)二維DOA估計(jì)分辨率低、精度低、算法復(fù)雜度高等問(wèn)題,本文提出基于MT?BCS算法的可分離二維DOA估計(jì)算法。該算法巧妙地將陣列流形矩陣A分解為俯仰維和方位維兩個(gè)獨(dú)立的低維導(dǎo)向矢量基矩陣,從而大大降低了算法的時(shí)間復(fù)雜度。而且算法對(duì)俯仰維、方位維進(jìn)行獨(dú)立估計(jì)大大提高了二維DOA估計(jì)的分辨率。由于不涉及對(duì)噪聲方差的估計(jì),算法的魯棒性也很高。在后續(xù)工作中將進(jìn)一步提高算法的分辨率,并降低其時(shí)間復(fù)雜度。
參考文獻(xiàn)
[1] GORODNITSKY I F, GEORGE J S, RAO B D. Neuromagnetic source imaging with FOCUSS: a recursive weighted minimum norm algorithm [J]. Electroencephalography and clinical neurophysiology, 2015, 95(4): 231?251.
[2] GORODNITSKY I F, RAO B D. Sparse signal reconstruction from limited data using FOCUSS: a re?weighted minimum norm algorithm [J]. IEEE transactions on signal processing, 1997, 45(3): 600?616.
[3] GORODNITSKY I F, RAO B D, GEORGE J. Source localization in magnetoencephalography using an iterative weighted minimum norm algorithm [C]// Proceedings of the 26th Asilomar Conference on Signals, Systems & Computers. Pacific grove: IEEE, 1992: 167?171.
[4] COTTER S F, RAO B D, ENGAN K, et al. Sparse solutions to linear inverse problems with multiple measurement vectors [J]. IEEE transactions on signal processing, 2005, 53(7): 2477?2488.
[5] MALIOUTOV D, ?ETIN M, WILLSKY A S. A sparse signal reconstruction perspective for source localization with sensor arrays [J]. IEEE transactions on signal processing, 2005, 53(8): 3010?3022.
[6] MALLAT S G, ZHANG Z. Matching pursuit with time?frequency dictionaries [J]. IEEE transactions on signal processing, 2013, 41(12): 3397?3415.
[7] DAVIS G, MALLAT S G, ZHANG Z. Adaptive time?frequency decompositions [J]. Optical engineering, 1994, 33(7): 2183?2191.
[8] PATI Y C, REZAIIFAR R, KRISHNAPRASAD P S. Orthogonal matching pursuit: recursive function approximation with applications to wavelet decomposition [C]// Proceedings of the 26th Asilomar Conference on Signals, Systems & Computers. Pacific grove: IEEE, 1993: 40?44.
[9] JI S, DUNSON D, CARIN L. Multitask compressive sensing [J]. IEEE transactions on signal processing, 2009, 57(1): 92?106.
[10] CARLIN M, ROCCA P, OLIVERI G, et al. Directions?of?arrival estimation through Bayesian compressive sensing strategies [J]. IEEE transactions on antennas & propagation, 2013, 61(7): 3828?3838.
[11] 劉自成.基于稀疏表示的雷達(dá)目標(biāo)角度與距離估計(jì)[D].西安:西安電子科技大學(xué),2014.
LIU Zicheng. Estimation of target′s angle and range in radar based on sparse representation [D]. Xian: Xidian University, 2014.
[12] Candès E J. Compressive sampling [C]// Proceedings of the International Congress of Mathematics. Madrid: European Mathematical Society, 2006: 1433?1452.
[13] 馬文潔.貝葉斯壓縮感知在DOA估計(jì)中的應(yīng)用研究[D].哈爾濱:哈爾濱工業(yè)大學(xué),2014.
MA Wenjie. DOA estimation through Bayesian compressive sensing algorithm [D]. Harbin: Harbin Institute of Technology, 2014.