非线性“超分辨”技术是谱分析领域里报导的新技术,这种技术在其他领域(其中包括射频(RF)自适应阵列天线系统)中同样也受到极大关注。本文选取了两项最流行的非线性技术—Burg的最大熵法和最大似然法。而非线性自适应阵列天线是由普通的旁瓣对消装置和增益定向约束技术(directional gain cons-traint techniques)组成的,与上述两项技术极为相似,本文将两者联系起来.所以,这基本上是一篇“互相补益”(cross-fertilization)的论文。本文以对比的方式进行了分析,这样可以使人们从天线空间方向图的观点验证它们的工作原理,也有助于对它们实际的超分辨性能进行鉴定。 本文归纳了几种多个非相参源自适应算法的分辨性能,并针对两个等强度源的情况,给出一个表示信噪比与以波束宽度为单位的源间距的关系的通用曲线图。自 适应算法对非等强度源也不难分辨,这是一个有意义的收获,本文对此也作了报导。用这些技术对相参的空间源或雷达目标进行超分辨比较困难,但每当有充分的相对运动或“多普勒周期”可资利用时,都得到了成功的结果。本文还提出了另外两种自适应空间谱的估值器,一个是预测其圆心的圆形阵列,另一个是新型的“热噪声”算法。 The non-linear “super-resolution” technique is a new technique reported in the field of spectral analysis, which is also of great interest in other fields, including the radio frequency (RF) adaptive array antenna system. This article selects the two most popular non-linear technology-Burg maximum entropy and maximum likelihood method. However, the nonlinear adaptive array antenna is composed of common sidelobe canceling devices and directional gain cons-traint techniques, which are very similar to the above two technologies, so this paper connects the two. Therefore, It’s basically an essay on cross-fertilization. This article analyzes in a comparative way that allows people to validate how they work from the perspective of the antenna’s spatial pattern and also helps to identify their actual super-resolution performance. In this paper, the resolving power of several non-coherent self-adaptive algorithms is summed up. And for two equal intensity sources, a generalized graph of the relationship between SNR and source spacing in beamwidth is given . Adaptive algorithms are also not difficult to distinguish between non-isosurfactor sources, which is a meaningful gain and is also reported in this paper. It is difficult to super-resolve these coherent spatial sources or radar targets with these techniques, but successful results are obtained whenever sufficient relative motion or “Doppler period” is available. This paper also presents two other estimators of adaptive spatial spectrum, one is to predict the circular array of its center, the other is a new “thermal noise” algorithm.