论文部分内容阅读
接收信号的谱估计是雷达系统在干扰环境中检测有用信号的最现代方法。必须指出,当相同距离单元中存在两种或多种杂波时(如地杂波和雨杂波),特别是当它们的多普勒频率未知时,这种处理是十分必要的。当前最普遍的方法是根据快速傅里叶算法(FFT)设计出一组滤波器(自适应或非自适应),这种方法称动目标检测(MTD)。这样可同时获得频谱分辨率和相干积累。为了克服由于谱失真引起的遮蔽效应,并为了获得更好的谱分辨率(由于抽样数据加窗而受到限止),根据另一种并不十分普遍的谱分析技术,探讨设计一种自适应滤波器的可能性。这些技术基于: (1) 选择一种适当的模式(通常选用自回归AR模式) (2) 模式参数的估计既可通过直接处理数据,也可通过间接地处理它们的相关函数来决定。在本文中,我们将把注意力集中在: (1) 估计误差引起的改善因子损失。 (2) 通过低计算成本技术获得相关函数的估计。 (3) 硬件分析以及性能和计算成本之间的折衷方法。
Spectral estimation of the received signal is the most modern way for radar systems to detect useful signals in an interference environment. It must be pointed out that this treatment is necessary when two or more clutters (such as clutter and rain clutter) exist in the same distance unit, especially when their Doppler frequencies are unknown. The most common method currently is to design a set of filters (adaptive or non-adaptive) based on the Fast Fourier Transform (FFT), a method called Moving Target Detection (MTD). This allows both spectral resolution and coherent accumulation. In order to overcome the shadowing effect due to spectral distortions and to obtain better spectral resolution (limited by windowing of the sampled data), another less common spectral analysis technique is explored to design an adaptive filtering Possibilities. These techniques are based on: (1) choosing an appropriate model (usually using an autoregressive AR model) (2) estimating the model parameters either directly by processing the data or by indirectly processing their correlation functions. In this article, we will focus on: (1) estimation error caused by loss of improvement factor. (2) Estimate the correlation function by low computational cost technique. (3) hardware analysis and compromise between performance and computing costs.