【摘要】：波束形成是陣列信號處理中的一項關鍵技術,在雷達、聲納、通信、醫(yī)療中有廣泛的應用。本文主要介紹了窄帶自適應波束形成的一些主要算法以及自適應陣列旁瓣相消和穩(wěn)健波束形成算法。在均勻線陣和窄帶信號模型條件下,給出了自適應波束形成的幾種算法,包括最大信噪比準則、最小均方誤差準則、最小方差無畸變準則,以及最小均方算法和采樣矩陣求逆算法。并且分析了各算法的原理和優(yōu)缺點,以及影響陣列自適應波束形成性能的兩個個主要因素,包括這列幅相誤差和自相關矩陣的估計誤差。理論上分析了自適應旁瓣相消系統(tǒng)的結(jié)構(gòu)和算法原理,基于最小均方誤差準則和采樣矩陣求逆算法,給出了自適應權(quán)矢量的求解方法,針對傳統(tǒng)旁瓣相消算法的缺陷又給出了另一種基于特征空間的算法,它能夠在小樣本情況下使旁瓣保持在一個較低水平。分析了造成波形不穩(wěn)健的主要原因,并介紹了幾種穩(wěn)健的波束形成算法,LCMV算法有較大損失,因為它展寬了主瓣,引入了更多的噪聲成份。對角加載法在SNR較大時性能有所下降,這是因為其固定的對角加載量無法隨著SNR的增加而變化,而且也無法確定一個合適的加載量。特征空間法在信噪比較低時性能損較大,這是因為在低SNR時,容易出現(xiàn)子空間纏繞,使得特征空間法失效。Worst-Case有比較好的性能,但是其不確定集參數(shù)的選擇仍然是一個問題。最后給出了一種基于二階錐規(guī)劃(Second-Order Cone Programming,SOCP)的迭代流程來近似原本的非凸優(yōu)化問題,其避免了使用等式約束,從而節(jié)省了系統(tǒng)自由度,進一步提升了系統(tǒng)性能。在理論分析的同時分別給出了各算法的仿真和性能分析。
[Abstract]:Beamforming is a key technology in array signal processing, which is widely used in radar, sonar, communication and medical treatment. This paper mainly introduces some main algorithms of narrowband adaptive beamforming and adaptive array sidelobe cancellation and robust beamforming algorithm. Under the condition of uniform linear array and narrowband signal model, several adaptive beamforming algorithms are presented, including the maximum signal-to-noise ratio criterion, the minimum mean square error criterion, and the minimum variance distortion-free criterion. And the minimum mean square algorithm and sampling matrix inverse algorithm. The principle, advantages and disadvantages of the algorithms and the two main factors affecting the performance of the array adaptive beamforming are analyzed, including the amplitude and phase errors and the estimation errors of the autocorrelation matrix. The structure and algorithm principle of adaptive sidelobe cancellation system are analyzed theoretically. Based on the minimum mean square error criterion and the inverse algorithm of sampling matrix, the method of solving adaptive weight vector is given. Aiming at the defects of the traditional sidelobe cancellation algorithm, another feature space-based algorithm is proposed, which can keep the sidelobe at a lower level in the case of small samples. This paper analyzes the main causes of waveform unsoundness, and introduces several robust beamforming algorithms, such as LCMV algorithm, which has a great loss, because it broadens the main lobe and introduces more noise components. The performance of diagonal loading method decreases when SNR is larger, because its fixed diagonal loading cannot change with the increase of SNR, and a suitable loading amount can not be determined. When the SNR is low, the performance of the feature space method is large, because it is easy to appear subspace winding in low SNR, which makes the feature space method invalid. Worst-Case has better performance, but the parameter selection of its uncertain set is still a problem. Finally, an iterative process based on second-order cone programming (Second-Order Cone Programming,SOCP) is proposed to approximate the original non-convex optimization problem, which avoids the use of equality constraints and thus saves the degree of freedom of the system and further improves the performance of the system. At the same time, the simulation and performance analysis of each algorithm are given.
【學位授予單位】：西安電子科技大學
【學位級別】：碩士
【學位授予年份】：2014
【分類號】：TN911.7

【相似文獻】

相關期刊論文 前10條

1 倪淑燕;程乃平;劉童嶺;;迭代更新導向矢量的穩(wěn)健波束形成算法[J];電訊技術;2009年11期

2 徐友根;劉志文;王四平;;二維正交矢量天線導向矢量的秩-1模糊問題研究[J];電路與系統(tǒng)學報;2006年02期

3 鄒翔;;導向矢量失配時的零陷加寬聯(lián)合穩(wěn)健算法[J];通信對抗;2014年01期

4 賈維敏;姚敏立;宋建社;;基于時空極化導向矢量的陣列信號多維參數(shù)聯(lián)合估計[J];信號處理;2007年06期

5 倪淑燕;程乃平;倪正中;;基于變對角加載的RCB算法[J];遙測遙控;2011年01期

6 冷文;王安國;;矢量修正解決DOA估計模糊問題[J];彈箭與制導學報;2010年03期

7 翁曉君;張e，

本文編號：2234275